ALASKA U.s. D FERC/DEIS-0038 ~~ /425 ,-- /"~ '::::J~-E , (/ pLr::rL VlO, t~5~ FEDERAL ENERGY REGULATORY COMMISSION OFFICE OF ELECTRIC POWER REGULATION

ARLIS Alaska Resources DRAFT ENVIRONMENTAL IMPACT STATEMENT Library & Information Services Anchorage, Alaska SUSITNA HYDROELECTRIC PROJECT FERC NO. 7114 - ALASKA

Volume 2. AppendixA. Load Growth Forecast: The Alaska Power Authority Forecasts Appendix B. Future Energy Resources . Appendix C. Energy Conservation Appendix D. 345-kV Transmission Line Electrical Environmental Effects

Applicant: Alaska Power Authority 333 West 4th Avenue Suite 31 Anchorage, Alaska 99501 I Additional copies of the Draft-EIS may be ordered from: Division of Public Information Federal Energy Regulatory Commission 825 North Capitol St., NE. Washington, D.C. 20426

May 1984 ~

; ; i

'"

CONTENTS ',' Page APPENDIX A. LOAD GROWTH FORECAST: THE ALASKA ~OWER AUTHORITY FORECASTS , A.l METHODOLOGY. , ...... ~.. ' A-3 A.2 LOAD PROJECTION ',1' .. A-4 A.3 WORLD OIL PRICE , ...... " A-4 A.3.1 Some Current Views ...... A-4 A.3.2 Masking Effect of Inventory Changes. A-5 A.3.3 Some Recent Trends and Their Meaning .' A-5 A.3.4 APA Oil Price and Load Projection . A-6 A. 3. 5 FERC Projections ...... A-13 REFERENCES ...... A-17

APPENDIX B. FUTURE ENERGY RESOURCES y B.1 INTRODUCTION ...... B-3 B.2 PETROLEUM FUELS...... B-3 B. J NATURAL GAS ...... B-3 B.3.1 Reserves/Resources .. • B-3 B.3.2 Pricing of Natural Gas . .. 8-4 B.3.3 Future Price of Natural Gas...... B-5 B.3.3.1 Completion of the ANGTS ...... B-5 B.3.3.2 Completion of Gas Pipelihe to Alaskan Gulf and Construction of LNG Export Facilities...... B-6 B.3.3.3 Construction of Facilities to Export Additional Volumes of Cook Inlet Gas ...... B-6 B.3.3.4 No Additional Facilities for Export of Cook Inlet Gas B-6 B.3.3.5 Future Gas Prices B-7 B.4 COAL ...... B-7 B.5 PEAT ...... B-B B.6 GEOTHERMAL ENERGY. B-B B.7 TIDAL POWER. .. B-B B.B SOLAR ENERGY .. B-8 REFERENCES ...... B-9 APPENDIX C. ENERGY CONSERVATION C.l ENERGY CONSERVATION AND THE NATIONAL ENERGY ACT OF 1978 ...... C-3 C.2 CONSERVATION OF OIL AND NATURAL GAS--THE POWERPLANT AND INDUSTRIAL FUEL USE ACT OF 197B ...... C-3 C.3 THE PUBLIC UTILITY REGULATORY POLICIES ACT OF 1978--RATE DESIGN, LOAD MANAGEMENT, AND REDUCTION OF THE GROWTH RATES IN THE DEMAND FOR ELECTRIC POWER C-4 C.4 RATE DESIGN AND LOAD MANAGEMENT--THE NARUC RESOLUTION NO.9 STUDY...... C-4 APPENDIX D. 345-kV TRANSMISSION LINE ELECTRICAL ENVIRONMENTAL EFFECTS D.l INTRODUCTION .... . D-3 D.2 OZONE PRODUCTION ...... D-6 D.3 AUDIBLE NOISE ...... D-B D.4 RADIO NOISE...... D-10 D.5 ELECTRIC AND MAGNETIC FIELDS D-19 D.5.1 Electric Fields D-19 D.5.2 Magnetic Fields D-21 D.6 ELECTRICAL SAFETY. D-22 REFERENCES ...... D-23 iv

LIST OF FIGURES Figure Page COVER PHOTO: Artist1s Rendition of the Proposed Watana Dam and Reservoir APPENDIX A. LOAD GROWTH FORECAST: THE ALASKA POWER AUTHORITY FORECASTS A-I Projected World Oil Prices ...... A-7 A-2 Price of Oil Under Various Forecasts ...... A-7 A-3 Alternative APA Load Projections for 1983-2010 ...... A-12 A-4 FERC Staff Load Projections and Selected APA Load Projections--1983-2010 A-16 APPENDIX D. 345-kV TRANSMISSION LINE ELECTRICAL EFFECTS D-l Susitna Project 345-kV Transmission System ...... D-4 D-2 Typical Tangent or Light-Angle Structure Placement Along Knik Arm-Gold Creek Section of Anchorage-Fairbanks 345-kV Transmission Corridor...... 0.-5 v ..

LIST OF TABLES Table Page APPENDIX A. LOAD GROWTH FORECAST: . THE ALASKA POWER AUTHORITY FORECASTS A-I APA's Reference Case World Oil Price Scenario . A-8 A-2 APA's Reference Case Railbelt Load Projection, 1983-2010 A-a A-3 APA's DRI "Base Case" World Oil Price Scenario, 1983-2005 A-9 A-4 APA's DRI "Base Case" Railbelt Load Projection, 1983-2010 A-:9 A-5 Implicit World Oil Price Scenario for·DOR Mean Projection A-I0 A-6 APA's DDR Mean Case Railbelt Load Projection, 1983-2010 ... A-I0 A-7 World Oil Price Scenario Implicit in DOR's 30% Case Projection A-l1 A-8 APA's OOR 30% Case Railbelt Load Projection, 1983-2010 .... A-l1 A-9 APA's Load Projections Relative to the Reference Case Forecast A-12 A-I0 Annual Load Growth Implied by APA Forecasts . A-13 A-II Railbelt Load Forecast, FERC High World Oil Price Scenario, 1983-2022 . A-14 A-12 Railbelt Load Forecast, FERC Medium World Oil Price Scenario, 1983-2022 A-14 A-13 Railbelt Load Forecasts of the Last Decade . A-17 A-14 Average Annual Expenditures for Electricity per Residential Household in the Railbelt . A-17 APPENDIX D. 345-kV TRANSMISSION LINE ELECTRICAL EFFECTS D-l Noise Levels of Typical Noise Sources . D-9 D-2 Calculated Audible Noise Levels for the Anchorage-Fairbanks Corridor with Three 345-kV Lines on a Common Right-of-Way Operating at 362.5 kV ... _ . D-I0 D-3 Summary of Noise Levels Identified by USEPA as Requisite to Protect Public ( Health and Welfa~e with an Adequate Margin of Safety . D-l1 D-4 Summary of Human Effects for Outdoor Day-Night Average Sound-Level of' 55 dB(A) . D-12 D-5 Audible Noise Complaint Guidelines Developed by Bonneville Power Administration . D-12 D-6 AM Radio Stations Received During Preconstruction Survey of Anchorage-Fairbanks Transmission Corridor between Willow and Healy, July 1981 . D-14 D-7 Existing Quality of Reception for AM Radio Stations . 0-14 D-8 Calculated Transmission Line Radio Frequency Noise Levels . D-16 D-9 Zones of Infl uence of Radi 0 Frequency Noi se ...... D-17 0-10 TV Stations Received During Preconstruction Survey of Corridor Route, July 1981 ...... D-18 0-11 Existing Quality of Television Reception . D-18 D-12 Possible EHV Line Effects on Communications Facilities and Recommended Mi nimum Cl earances ...... D-20 D-13 Calculated Intertie Electric Field Strengths . D-22 0-14 Right-of-Way Use of Single and Multiple Single-Circuit Transmission Lines D-23

ARLIS Alaska Resources Library & Information serVices Anchorage, Alaska "" DRAFT ENVIRONMENTAL IMPACT STATEMENT SUSITNA HYDROELECTRIC PROJECT, FERC NO. 7114 .,(-

APPENDIX A LOAD GROWTH FORECAST: THE ALASKA POWER AUTHORITY FORECASTS

prepared by Federal Energy Regulatory Commission Staff

A-I A-3

~

APPENDIX A. LOAD GROWTH FORECAST: THE ALASKA POWER AUTHORITY FORECASTS

A.1 METHODOLOGY The Applicant has submitted a number of alternative load forecasts for the Railbelt, based on varying world oil price scenarios. All these forecasts were generated by means of the same modeling structure. That structure employs three computer-operated models that provide projec­ tions of: (1) regional demographic, and state economic and fiscal variables, (2) regional electricity demands given specific energy price assumptions, and (3) least-cost generation expansion programs given a demand forecast. The last two model s are iterated to determine a consistent electricity demand forecast given the cost of power projected by the generation expansion program appropriate to that demand forecast. . The first computer model--the Man in the Arctic Program (MAP) Economic Mbdel--operates for each of 20 regions within the state; the Railbelt consists of six of those regions. Region-specific projections are produced by disaggregating a statewide projection of employment, population, and household formation variables. The state-level economic, fiscal, and population portions of MAP are solved algebraically in simultaneous fashion. That is, equations within the economic portion of the model are dependent, for instance, on projections within the population portion of the model. The population projections also are dependent on the economic projections. This inter­ dependence, or simultaneity, requires the MAP model to solve iteratively for each year1s set of projections. The fourth portion of MAp--the household formation portion--is not interdependent with any of the other projections, but merely produces projections based on the results of the population forecast. While the many simultaneous and recursive relationships, as well as the large number of equations (more than 1,000) contained in MAP, suggest a highly complex forecasting system (which it is), it is also the case that a great deal of critical information concerning the Railbelt economy has to be forecast exogeneous to the MAP model. For instance, employment projections for the most important sectors of the basic economy have to be assumed. Similarly, large components of the state's projected revenues--a dominant influence in the Railbelt economy--have to be assumed in order to generate forecasts with MAP. The inability of MAP to generate projections for some of these economic variables is due in part to their dependence on influences outside the economy of Alaska. (For instance, employment within the fishing industry is determined in the main by demand fOI' Alaska IS fi sh products in the export markets.) In other instances, independent mode 1i ng efforts conducted by unaffi 1i ated organi zations have been used to formul ate assumed values for some of the MAP data inputs. The MAP projections rely, for instance, on some fore­ casted data prepared by the Alaska Department of Revenue. The MAP model operates to produce annual forecasts through the year 2010. Output from the MAP model that is used subsequently by the Rai'lbelt Electricity Demand (RED) Model as input data consists of annual population projections by load center, total annual employment by load center, and annual household formation projections by load center. The RED model requires exogeneous forecasts of retail pri ces for fue 1 oi 1, natural gas, and e1ectricity. The projections of electricity demand produced by the RED model are customer-class-specific for three categories of customers--residential, business and miscellaneous--and represent total annual kilowatt-hour (kWh) consumption at the customer's meter for five-year intervals. Linear interpolation of these forecasts is used to derive annual projections. The residential consumption portion of RED employs an end-use approach that recognizes nine major end uses and one catch-all category of end use appropriate to this group of consumers. The total stock of electricity-consuming appliances and equipment is a function of time and the type-of-household-formation projections generated with RED. (The latter are consistent with MAP model input data for households.) Vintage-specific electricity consumption profi les for the various end uses are combined with the stock projections to compute energy usage before making adjustments for fuel price changes. The price-induced consumption adjustments are premised on assumed values for Railbelt own- and cross-price elasticities associated with electricity, natural gas, and fuel oil prices. The business consumption portion of the RED model actually encompasses the commercial, small industrial, and government sectors of the Railbelt. Aggregate electricity consumption in the absence of any change in fuel prices is forecast as a function of regional commercial floor space, which is derived from an ad hoc assumption regarding future trends in the relationship between floor space and total emplOyment. The price-induced changes in consumption of electri­ city by the business sector are modeled in a fashion similar to that used in the residential sector. That is, the values for own- and cross-price elasticity terms are assumed. A-4

The mi see11 aneous sector e1ectrici ty consumpti on projections represent use for street 1ighti ng, vacation homes, and vacant dwellings. These consumption projections are forecast by assuming a multiplier for total residential and business sector kWh consumption representative of street lighting requirements, a multiplier for total number of households times a constant kWh con­ sumption factor to represent vacation home electrical consumption, and a multiplier for the total number of vacant houses representing vacant-dwell i ng kWh consumption. The sum of these three products is the projected miscellaneous-section consumption. In"addition to the residential, business, and miscellaneous sectors, a fourth component of elec­ tricity consumption is appended to each year's kWh projection. This component is identified as "exogenous i ndustri all oad. " The kWh load projected for this customer category is an ad hoc forecast based on the judgment of a consulting firm that participated in the preparationo{the license application. The Applicant's projections of annual peak demand within the Railbelt are computed by means of a load factor multiplier that operates on the kWh projections to produce the peak kW demand. Load factor is defined as the ratio between the average hourly kW demand for the year and the annual peak kW demand for the year. Thus, dividing the annual kWh load projection from a RED model forecast by the number of hours in the year (i.e., 8,760 in a non-leap year) and then dividing by the load factor results in a figure for peak demand. The load factors used in the Applicant's projections are assumed values specific to the Anchorage area and to the greater Fairbanks area of the Railbe1t. These assumed load factors are the simple averages for the period 1971-1980 for each of the two regions. The project variation over time in the implied load factor for the Ra ilbe 1t as a whole deri ves from the varyi ngcontributions to total kWh load attri butab1e to Anchorage and Fairbanks over the forecast period. A.2 LOAD PROJECTION The Applicant has prepared load projections for 1983-2010 under a wide range of alternative scenarios. Each forecast scenario is characterized by a specific trajectory for the price that crude oil will command in world markets over the forecast horizon. There are at least three reasons that the world oil price is chosen as the single exogenous variable to be altered in attempting to bracket the load growth in the Rail belt. First, world oil prices affect the level of petroleum revenues to the State of Alaska, mainly through sever­ ance taxes and royalty payments. These revenues account for more than 80% of total state revenues, and the 'state is the single largest economic force acting on the Rai1belt economy. Second, world oil prices affect directly the costs of electricity generated in the Railbelt because of the linkage between prices of crude and other fossil fuels. As demonstrated in Section 1.2, the Rai1belt depends heavily on fossil-fired electric generation. Third, world oil prices, through their influence on other fuel prices, affect the substitution possibilities that exist for electricity in the Railbelt. A.3 WORLD OIL PRICE A.3.1 Some Current Views There is little consensus in views concerning future world oil prices. Oil price forecasts for the year 2010 range from as low as about $12 per barrel to $110 per barrel ($88 to $809 per metric ton) (in 1983 dollars). Clearly, there is considerable uncertainty concerning future oil prices. The uncertainty can be traced back to one fact--since late 1973, the price of oil has contained a large element of monopoly profit.* The high oil price projections are all based on an inherent assumption that the OPEC nations will maintain their market power and continue to extract large monopoly profits from the price of oi1.** The lower oil price projections derive from an inherent assumption that the OPEC nations will lose much of their market power and that prices will fall toward the marginal cost of finding and producing new oil. The OPEC nations already have lost most of the market power they possessed before 1979. The rapid decline in

*As used here, monopoly profit is defined as the difference between the actual price of oil and the price it ,would bring in- a fully competitive market. For example, assume that the actual price of a .barre1 of oil is $29/barrel ($213/metric ton), that it would only be $15/barre1 ($110/metricc ton) in a fully competitive market (its cost to the marginal cost producer), and that the cost to actually produce a barrel of Middle East crude oil is only $3/barrel ($22/ metric ton). ~Middle Eastern country would thus extract $14 ($103) monopoly profit and $12 ($88) economic rent (or producer surplus) from each barrel produced and sold. Its total profit (economic rent and monopoly profit)"'would be $26/barrel ($195/metric ton). **Market power is possessed whenever a group of producers, by restricting production, are able to maintain the price of a product higher than it would otherwise be i'n a fully competitive market. Market power is a requirement for extracting monopoly profits. A-5

OPEC oil demand (from 31 million barrels per day [mmb/d] [4.2 million metric tons per day (MT/d)] in 1979 to 14.3 mmb/d [1.9 million MT/d] in February 1983) forced these nations to reduce the price of oil during March 1983. A further decline in OPEC ~ildemand would likely cause further price cuts. Consequently, the key question in predi cti ng future oil prices relates to whether, demand for OPEC oil will remain strong enough to allow the OPEC nations to continue to extrcft;t monopoly profits from the price of oil. If so, how much can they extract and for how 10ng?'Ylf not, then how far will prices fall? Those individuals forecasting higher oil 'prices assume that a strong upturn in the world economy will increase world oil consumption and cause incl'eased prices. Forecasters projecting lower oil prices assume that the demand for OPEC oil will continue to fall in spite of an improving world economy due to continuing fuel switching, conservation, and a growth in non-OPEC oil production, causing a loss of market power and a further oil price decline. Most forecasters expecting oil prices to rise acknowledge that if the demand for OPEC oil continues to decline, then oil prices also wi 11 fall. The differences in oil price forecasts, therefore, stem from di fferent expectations of future demand for OPEC oil.' A.3.2 Masking Effect of Inventory Changes Oil prices have stabilized and OPEC oil production has increased since the March 1983 oil price reduction. While OPEC oil production has risen, oil consumption likely has continued to fall. This discrepancy between production and consumption results from inventory changes. Throughout most of 1982, world petroleum inventories were reduced by about 1.5 mmb/d (200,000 MT/d). Thus, 1982 oil consumption was actually higher than indicated by production data. In addition, about 200 million barrels (27 million MT) were withdrawn from storage immediately prior to the March 1983 official oil price reduction. This abnormal inventory drawdown resulted in about a 4.5 mmb/d (600,000 MT/d) reduction in OPEC oil demand during a season when demanu normally increases. Actual oil consumption during the period was several million barrels per day higher than oil production. Recent OPEC oil production levels of 17.5 to 18.5 mmb/d (2.4 to 2.5 million MT/d) should not be viewed from the perspective of OPEC's February 1983 production level [14.3 mmbld (1.9 million MT/d)] , which was abnormally low due to rapid inventory withdrawals, but from the perspective of the approximately 20 mmbld (2.7 mi 11 ion MT/d) average rate that OPEC woul d have produced duri ng 1982 had it not been for inventory drawdowns. OPEC's recent production is about equal to its expected oil demand, assuming world oil consumption has continued to decline relative to energy consumption as it did in 1982. Thus, the true demand for OPEC oil still appears to be declining. If so, then OPEC may have difficUlty maintaining the current oil price structure. A.3.3 Some Recent Trends and Their Meaning* Spot-market oil prices have declined approximately 27% (in nominal dollars) since they peaked in 1981. There is considerable speculation that they may fall again soon. Thus, the OPEC nations have lost much (but not yet all) of their market power. Oil has rapidly lost its share of the world's energy consumption. It lost a 6% share during the last three years. The free world's oil production declined 10.lmmb/d (1.4 million MT/d) from 1979 through 1982. Adjustments for inventory changes indicate that oil consumption declined 7.2 mmb/d (980,000 MT/d). Of this, 5.7 mmb/d (775,000 MT/d) (79%) resulted from a reduction in oil's share of total energy consump­ tion. Oi 1 production has decl ined 7% per year during the recent world economic I'ecession compared with a decline of only 2% in total energy production (5% and 1% when adjust­ ments are made for inventory changes). The rapid loss in market share indicates that oil is currently overpriced relative to other fuels. Oil's share of the world's energy consumption was declining slightly, prior to 1979 [when its price was around $17.60 per barrel ($129/MT) expressed in 1983 dollars]. The price at which oil would not lose market share may be as low as $14 per barrel ($103/MT) , but likely is somewhat higher. Conservation has reduced world energy consumption per unit of ecoQomic output. Since 1979, world energy consumption per unit of the world's Gross Domestic Product (GDP) has declined at a rate of about 2% per year. Prior to 1973 the growth in the world's energy consumption was about equal to the growth in the world's GDP. From 1974 through 1979 it fell below growth in the GDP by about 1%. The statistical evidence currently available does not indicate that the rate of conservation is declining.

*All statistics and analyses in this section are based on Essley, 1983.

~ A-6

Non-OPEC oil production has risen 6 mmb/d since 1976 at a compound growth rate of 5.3% per year. It increased 5% during the first six months of 1983 compared with 1982. Un 1ess oil prices fa 11 further, the 1arge profit from oi 1 production shoul d continue to draw large capital funds for exploration and development in non-OPEC countries.

OPEC oil production may continue to fall. It dropped 12 mmb/d (1.6 million MT/d) during 1980 through 1982. If the world economic recovery is weak, if fuel switching and conservation continue at near their recent rates, and if non-OPEC oil production continues to rise, then OPEC oil production could decline another 3 to 7 mmb/d (400,000 to 950,000 MT/d) during 1983 and 1984. Even a cursory analysis of recent trends indicates that oil prices could decline further. Some analysts believe that market forces affecting oil prices will be so strong that it is only a question of "when" prices fall rather than "if" they fall. Of course, mil itary conflict could disrupt oil supplies and even cause an increase in oil prices. However, any supply disruption and subsequent price increase would be temporary. Once supplies were restored, the same forces currently tending to cause oil prices to fall, but amplified by the supply disruption and higher prices, would again exert a strong pressure for a lower oil price. To assume that OPEC oil production will increase in the near term to the extent that some ana­ lysts have projected requires assumptions of: Strong worl d economi c recovery and future growth (i. e., hi gher than most economi sts are generally projecting), Reduced fuel switching, Reduced energy conservation, and A leveling or decline in non-OPEC oil production. Such events are possible, and some analysts projecting increased oil prices (or even stable oil prices) obviously consider them more likely than a continuation of recent trends. Nevertheless, almost all analysts agree that there is so much uncertainty that any oil price projection, whether up or down, should be viewed with circumspection.

Figure A-I shows the oi 1 price range that FERC cons i ders to be most 1ikely. FERC 's mi d-range projections, expressed in 1983 dollars, are as follows: Year 1983 1985 1990 1995 2000 2010 Oil price ($/barrel)* 29 24 20 22 24 29

FERC's projection is based on an assumption that the strength of economic forces now acting in the direction of reducing oil prices (fuel switching, conservation, and the growth of non-OPEC oil production) will continue to exceed the strength of economic forces tending to increase oil prices (renewed world economic growth). Figure A-2 shows several oil price projections by Alaska's Department of Revenue; Sherman H. Clark, Associates (SHCA), consultants to the Alaskan Power Authority; and DOE. The SHCA and DOE projections are all postulated on an assumption that the combination of economic forces will cause a sufficient growth in demand for oil to allow OPEC to incr~ase its output, and hence maintain its market power. If oil prices decline, then the magnitude of fuel switching and conservation should diminish, less exploration and development should occur in non-OPEC countries, and the world's economic growth should be stimulated. In short, a reduction in oil prices will reduce the magnitude of forces tending to further reduce oil prices and will increase the magnitude of forces tending to cause prices to rise. As a consequence, even if oil prices dec 1i ne in the near term, they eventually will start to rise again. Almost all analysts project increasing prices after about a decade or less. Conversely, if oil prices rise, then the economic forces tending to cause oil prices to fall will be strengthened, whereas the degree of the world's economic recovery will tend to ~e'reduced. A.3.4 APA Oil Price and Load Projection The APA takes asits reference case for the worl d oil price scenario a projection made by Sherman H. Clark Associates, a CaJifornia-based energy consulting firm. The forecasters respon­ sible for this oil price projection have assigned a 35% probability of occurrence to this

*$l/bbl = $7.35/metric ton. --I

A-7

"" 100 ,..--- I

~,' -~~~, " +~, " ," , Possible Supply Disruption "," - ~~' \ -~~ .II ~~ II ~~~~ " ,~.".~-~ \ ...~~~-~~ \'* _---- __------_I Minimum--- ,V----,,------Competitive Pricing Potential -- Possible Decline Due---to (price collapse) Photovoltaics, Fusion, etc.

1970 1980 1990 2000 2010 2020 2030 2040 2050 Year Figure A-I. Projected World Oil Prices.

(1) 100

/(2) 1. DO ENNEP-83 Scenario C 2. Sherman H. Clark Associates-Supply 80 / Disruption /" 3. DOE NNEP-83 Scenario A (low case) CI:I " ... / 4. Sherman H. Clark Associates-No Supply S "E /" Disruption - '1:S /. 5. Alaska Department of Revenue-Mean N 60 co /. /(3) 6. Alaska Department of Revenue- 50% en / 7. Alaska Department of Revenue-30% ...... /. , •• (4) I / "~ ..- is ~. ~..... 40 .~. ,-~­ -c u,CD ; / __ (5) -;:: . ~ .",.,...... _., ~ . , -.~-'.. --"- } 20 ~.... ·-··_··_••_ ••_(6 ------(7}

1970 1980 1990 2000 2010 2020 Year

Figure A-2. Price of Oil Under Various Forecasts. A-8 particular scenario. Among other things, this forecast, according to APA, assumes "that OPEC will continue operating as a viable entity and will not limit production during the forecasted period. Recent trends in economic growth in the United States and the free world ~ill continue at reasonable rates." The particular prices for world crude associated with this reference case are shown in Table A-I. State petroleum revenues consistent with this world oil price trajectory are computed and are input to the MAP model to begin the load forecasting sequence. The results of that forecast procedure are shown in Table A-2.

Table A-I. APA's Reference Case World Oil Price Scenario

Price in Annual Rate Final Year of of Change in Period Price Year(s) (1983$/bbl) (%) 1983 28.95 -14.9 1984 27.61 -4.7 1985-1988 26.30 -1. 2 1989-2010 50.39 2.6 Conversion: $1/ bbl= $7. 35/metric ton Source: Based on data from Applic~tion Volume 2a.

Table A-2. APA's Reference Case Railbelt Load Projection, 1983-2010

Energy Peak Demand Year (Gwh) (MW) 1983 2,803 579 1985 3,096 639 1990 3,737 777 1995 4,171 868 2000 4,542 945 2005 5,093 1,059 2010 5,858 1,217 Source: Based on data from Application Volume 2C.

Load projE!ct ions also are made that use the "base case" forecast of worl d oil price constructed by Data .~Resources, Inc. (ORO. Accordi ng to APA, the DRI forecast makes ass umpt ions si mil ar to the Sherman Clark projections regarding the continued influence that OPEC will yield on world oil markets, as well as the economic growth to be exhibited by the U.S. economy. DRI's forecast of world oil prices are, however, noticeably different than the reference case scenario, as is shown in Table A-3. State petroleum revenue inputs to the MAP model are prepared similarly to the procedure described above, and the load forecasts result is shown in Table A-4. ---l

A-9

Table A-3. APA's ORI IIBase Case ll World Oil Price Scenario, 1983-2005 ....

Price in Annual Rate Final Year of of Change) n Period Pri ce,,- Year(s) (1983$/bb1) (%) 1983 28.95 -14.9 1984 25.17 -13.1 1990 36.99 6.6 2000 53.43 3.7 2001-2005 56.54 1.1 Conversion: $l/bbl = $7. 35/metric ton

Table A-4. APA's DRI IIBase Case ll Railbelt Load Projection, 1983-2010

Energy Peak Demand Year (Gwh) (MW) 1983 2,811 580 1985 3,109 642 1990 3,717 773 1995 4,341 904 2000 5,041 1,050 2005 5,857 1,220 2010 6,965 1,450

A third load projection is presented by APA that is premised on the Alaska Department of Revenue's (DOR) mean probability estimates of state petroleum revenues. These petroleum revenue figures are translated into implicit prices for world oil that are presented in Table A~5. The load projections associated with this trajectory of world oil prices are summarized in Table A-6. A fourth projection, which repeats the process just described for the DOR's mean probability case, is made using the DOR's "30% probability" case. That is, the Department of Revenue has forecast a level of state petroleum revenues that their model projects has a 30% chance or less of not being exceeded. (A more straightforward way of interpreting this is that there is a 70% chance that state petroleum revenues will exceed the amount forecast under this case.) Again, the implicit world oil prices consistent with these petroleum revenue projections are derived, and the inputs to MAP are calculated. The world oil price trajectory and associated load fore­ cast for this scenario are shown in Tables A-7 and A-8. For comparison, all four of these alternative load projections are depicted in Figure A-3.

ll Us i ng APA IS"Reference case as a standard for compari son, it shoul d be noted that there is little to distinguish these projections in the near term. Variation around that Reference case load projection is less than 3.5% in 1985, as shown in Table A-9. By 1990, however, significant differences exist in the forecasts. Implied annual growth rate in kWh loads during that period range from a high of 3.8% in the Reference case to a low of 2.2% in the DOR 30% scenario, as shown in Table A-10.

~ A-10

Table A-5. Implicit World Oil Price Scenario for DOR Mean Projection

Price in Annual Rate Final Year of of Change in Period Price Year(s) (l983$/bbl) (%) 1983 28.95 -14.9 1984 23.96 -17.2 1985 22.67 -5.4 1986 22.35 -1.4 1987 21. 95 -1. 8 1988-1999 25.60 1.3 Conyers ion: $1/bbl =$7.35/metric ton

Table A-6. APA's DOR Mean Case Railbelt Load Projection, 1983-2010

Energy Peak Demand Year (Gwh) (MW) 1983 2,776 573 1985 3,050 630 1990 3,508 730 1995 3,849 801 2000 4,228 879 2005 4,726 982 2010 5,399 1,121 A-ll

'"

;-; Table A-7. World Oil Price Scenario Implicit in DOR's 30% Case Projection·

Price in Annual Rate Final Year of of Change in Period Price Year(s) (1983$/bb1) (%) 1983 28.95 -14.9 1984 22.74 -21. 5 1985 21. 00 -7.7 1986 20.32 -3.2 1987 19.52 -3.9 1988-1999 14.76 -2.3 Conversion: $l/bbl = $7.35/metric ton

Table A-8. APA's DOR 30% Case Railbelt Load Projection, 1983-2010

Energy Peak Demand Year (Gwh) (MW) 1983 2,753 568 1985 3,014 622 1990 3,364 699 1995 3,560 740 2000 3,890 808 2005 4,343 926 2010 4,950 1,026 A-12

GW~

7000 DRI

6000 Reference

DDR mean

5000 DDR 30%

4000

2800

1985 1990 1995 ·2000 2005. 2010

Figure A-3. Alternative APA Load Projections for 1983-2010.

Table A-9. APA's Load Projections Relative to the Reference Case Forecast

Forecast Scenario 1985 1990 1995 2000 2010 DRI 1. 00 0.99 1. 04 1.11 1.19 Reference 1. 00 1.00 1. 00 1.00 1. 00 DOR Mean 0.99 0.94 0.92 0.93 0.92 >OOR 30% 0.97 0.90 0.85 0.86 0.85 -l

A-13

Table A-10. Annual Load Growth Implied by APA Forecasts (percent) '"

Forecast Scenario 1985-1990 1990-1995 1995-2000 2000-2010 ORI 3.64 3.15 3.04 3;£9 Reference 3.84 2.22 1.72 2.60 DOR Mean 2.84 1. 87 3.80 2.47 DOR 30 2.22 1.14 1. 79 2.44

By 1995, the differences are more pronounced. The DRI base case scenario is higher than the Reference case by 4%, and the OOR 30% case is lower by some 15%. The average annual growth in kWh load implied by the high and low cases in 1995 varies by more than 175%. Between 1995 and 2000, the APA Reference case exhibits the lowest average annual load growth of any of the scenarios, despite being 7% and 14% higher in absolute terms than the DOR mean and 30% cases, respectively. The ORI-based scenario continues to exhibit better than 3% annual growth during this period. There is little change in the relationships among these alternative forecasts during the period 2000-2010. The Reference and both the DOR cases have converged on average annual growth rates of near 2.5%. The DRI-based increases its growth sl ightly to 3.3% per year. An important implication in the relationships among these scenarios is the insulation exhibited between e1ectricity load growth and worl d oil pri ces in the event that those pri ces are assumed to decl i ne. Note that the Reference case is characterized by worl d oil prices that grow at 2.6% annually in real terms during this interval. The DOR 30% case has world oil prices that decline throughout the period. Under both scenarios, however, electricity growth is virtually the same during the ten-year period. The reasons for this behavior in the model forecasts have been analyzed by the APA and are discussed below. A.3.5 FERC Projections The FERC has judged the world oil price trajectories described earlier to be more plausible than the oil price scenarios recommended by the Applicant. As a consequence, an additional series of load projections have been made using these world oil price forecasts. The projections use the same modeling apparatus constructed by the APA and require conversion of the world oil price forecast to a forecast of state petroleum revenues for use in the MAP model.* This conversion was carried out in a manner consistent with the one used by APA. Further, the RED model input requirements for end-user fuel prices were made consistent with FERC world oil price trajecto­ ries. The load projections that resulted for the medium and high world oil price assumptions are shown in Tables A-II and A-12. No projections consistent with the low world oil price trajectory could be generated. The state economic model component of MAP was unable to compute a solution given the drastic reductions in state revenues implied by the low oil price in 1985. This should not be viewed as a failure of the MAP model. The result is indicative of the very serious economic problems the world and Alaska, in particular, are likely to face if the price of oil collapses to the $10 barrel range in 1985.

*It should be noted that in addition to the changes in world oil price scenarios that FERC chose to make, alterations to the MAP model also were pursued. The objective in making these alter­ ations was to improve what FERC judged to be the economic consistency of what appears to be a sophisticated forecasting tool. Nevertheless, where the specification of an equation could be altered to add economic content, as well as improve both the statistical fit and significance of coefficients in the equation, then such a modification was made. In those instances when an equation was successfully altered, it was also the case that substitution of the new equation into the model caused the system to become unstable. This was the case because critical link­ ages within the system of equations were broken as a consequence of the changes made by FERC. This can occur despite the changes' having improved the particular equation viewed in isolation. This is not an unreasonable circumstance given a model with the complexity of the MAP system. For this reason, FERC has jUdged that the forecasting models employed by the Applicant could not be improved on in the time allotted, and these same models have been adopted for purposes of generating the FERC Railbelt forecasts. A-14

Table A-11. Railbelt Load Forecast, FERC High World Oil Price Scenario, 1983-2022

Energy Peak Demand Year (Gwh) (MW) 1983 2,814 581 1985 3,116 644 1990 3,567 742 1995 3,927 817 2000 4,447 925 2005 4,793 996 2010 5,371 1,115 2020 6,591 1,367 2022 6,866 1,424

Table A-12. Railbelt Load Forecast, FERC Medium World Oil Price Scenario, 1983-2022

Energy Peak Demand Year (Gwh) (MW) 1983 2,802 579 1985 3,094 639 1990 3,474 722 1995 3,788 788 2000 4,168 866 2005 4,623 960 2010 5,234 1,086 2020 6,424 1,332 2022 6,693 1,388 A-15

There are a number of ways to put these alternative projections into perspective. Three approaches are used here. The first is a simple comparison of the FERC projections and the APA projections. The second is a comparison of both sets of'forecasts with other projections made previously for the Railbelt. The third is an examination of the changes in both electricity intensity and electricity expenditure implied by these forecasts. A graphical comparison of APA and FERC projections is shown in Figure A-4. It i~~

GWh 6000 APA reference

5500 OOR mean FERC high FERC medium 5000 OOR 30%

4500

4000

3500

3000

1985 1990 1995 2000 2005 2010

Figure A-4. FERC Staff Load Projections and Selected APA Load Projections--1983-2010. 1

A-l7

There are significant differences in the implied expenditures for electricity across the various load forecasts. Average electricity expenditures per household are shown in Table A-14 for a representative of the projections. ~ The variation in the real per household expenditures for electricity should presumably be reflec­ ted in the usage intensity figures discussed previously.

Table A-13. Railbelt Load Forecasts of the Last Decade (GWh and MW)

Forecast 1980 1985 1990 1995 2000 H. J.Kaiser Co. 3,543 6,434 11,701 1974 (677)tl (1,194) (2,155) Railbelt Utilities 3;514 -- 10,377 1974 (537) (1,560) Corps of Engineers 3,240 -- 6,840 11,650 1975 (740) (1,480) (2,660) Alaska Power Administration 3,155 -- 6,110 10,940 1976 ISER 2,790 -- 4,030 5,170 6,430 1980 (510) (735) (934) (1,175) Battelle -- -- 4,456 4,922 5,469 1981 (892) (983) (1,084) Battelle -- -- 4,482 4,894 4,728 1982 Battelle -- 3,096 3,737 4,171 4,542 . 1983 FERC -- 3,094 3,474 3,788 4,168 t 1 Numbers in parentheses are peak load forecast.

Table A-14. Average Annual Expenditures for Electricity per Residential Household in the Railbelt (1982 $)

Forecast 1985 1995 2005 APA Reference Case 713 834 955 FERC Medium Case 712 774 869 DOR 30% Case 709 706 706

REFERENCE FOR APPENDIX A Ess 1ey, P. L., Jr. 1983. Future Worl d Oil Prices--Wi 11 They Ri se or Fa 11? ORA/FERC. (November 21). ,

DRAFT ENVIRONMENTAL IMPACT STATEMENT .' SUSITNA HYDROELECTRIC PROJECT, FERC NO. 7114

APPENDIX 8 FUTURE ENERGY RESOURCES

Prepared by Federal Energy Regulatory Commission Staff

8-1 B-3

"".

APPENDIX B. FUTURE ENERGY RESOURCES B.l INTRODUCTION The physical availability of energy resources in the Railbelt is not a significant issue. There are sufficient reserves of oil, gas, and coal,each taken individually, to meet the most opti­ mistic projections of internal Railbelt energy use from now until well past the mid-21st Century. The cost at which any of these resources will be made availab"le to Railbelt consumers is what is at issue, and it is the means of measuring such cost that is central to this discussion. The real cost of consuming an energy resource within the Railbelt is not necessarily the sum of the labor, capital, materials, and assorted other production expenses required to extract and convert the resource to usable energy. The real cost is what that resource will command in the market that values it most" That market can, and often does, lie outside the region. Where the export price (i.e., value) exceeds the cost of producing the resource for local consumption, it is the export markets' payment foregone that is the cost of consuming the resource locally. By consumi ng the energy locally, in thi s instance, the opportuni ty to recei ve the export value is lost. Therefore, depending on what assumptions are made about the "highest valued use" to which the Railbelt's energy resources can be put, there will be radically different circumstances that characterize both the economic availability of energy resources and the most efficient means of meeting the energy requirements of the region. Although the Railbelt has been able to meet its current energy requirements at reasonable cost (in certain cases, at comparatively modest cost), the issue of future energy cost is subject to debate. The major source of controversy stems from the future course of world oil prices and their relationship to the supply and demand for fossil fuels of all types. For this reason a di scussi on of the worl d oi 1 pri ce and its probable future range is necessary., B.2 PETROLEUM FUELS The supply of petroleum fuel is related to the supply of crude6il. From Alaska's point of view, its crude oil reserves are so large relative to its internal needs that supply should not be a constraint on the use of petroleum fuels for the foreseeable future. Price i5 another matter. If oi 1 prices ri se re1ati ve to other energy resources, whi ch al so are abundant in Alaska, then the state may receive the greatest economic benefit from "exporting". its petroleum resources while consuming its lower-cost resources. Petroleum fuel consumption could become "demand-constrained". ' The prices of petroleum fuels obviously are related to the price of crude oil. However, refin­ ing costs are independent of the price of crude oil; hence, fluctuations in crude oil prices will not cause similar fluctuations in .refining costs. The cost of refining is unlikely to change appreciably over time (in real dollars). As a consequence, variation in crude oil prices should result in equivalent dollar variations in petroleum fuel prices (but not equivalent percentage changes). Various petroleum fuels'exceed crude costs by different amounts, due to different refining costs and differences i ndemand. Normally, gaso 1i ne has the hi ghest refi nery markup of the hi gh­ volume petroleum fuels. Residual fuels normally are by-products and generally are sold for less than the cost of crude oil. Recently, however, high-sulfur residual fuels have commanded a higher price than low-sulfur crude oil on the U.S. West Coast and in Alaska. This abnormal condition appears to be a result of the peculiarity of U.S. export laws (Tussing, 1983)." Projec­ tions of future residual fuel prices thus are even more uncertain than projections of crude oil prices, since an additional political uncertainty is added. B.3 NATURAL GAS 8.3.1 Reserves/Resources Alaska's proven gas reserves far exceed its internal needs for the foreseeable future under even the most extremely optimistic projections of growth. Further, its potential gas resources may materially exceed its proven reserves. The amount of gas required to generate all the Rail­ belt's electric power needs for the next half century [about 3 trillion cubic feet (Tcf)] is 1i ke ly 1ess than 10% of Alas ka I s proven gas reserves and perhaps 4% of its potential gas resources. Paradoxically, some have suggested that Alaska's gas reserves may not be sufficient to prudently plan to use gas for future electric power generation. 8-4

This paradox results from the location of Alaska's gas reserves and its potential gas resources. The bulk of Alaska's gas may not be accessible for use to generate power in the Railbelt area, may be accessible only after it is needed, or may be accessible only at a cost that prohibits its use. Unless oil prices increase materially, a pipeline to transport Prudhoe Bay gas may not be constructed. If oil prices follow FERC's projections, for instance, Prudhoe Bay gas may remain locked in place well into the next century. The Cook Inlet proven reserves, while readily accessible to the lower Railbelt area, may not be sufficient to meet the area's power needs for more than about 20 years if consumption continues at the present rate. However, in addition to the approximately 3.4 Tcf of proven reserves in the Cook Inlet area, the United States Geological Survey indicates that there is likely another 1.3 to 13 Tcf of gas as yet undiscovered. If so, then there should be more than adequate gas to meet the Railbelt's power needs for the next half century. But since such potential reserves are not proven, and may not materialize, it is argued that it would be imprudent to plan on the use of the as yet undiscovered gas. Further, it is argued that even if the gas is present, gas prices will have to rise materially to ensure that it is discovered and developed. If Prudhoe Bay gas reserves remain locked in place, and if no new reserves are discovered in the Cook Inlet area, then a strategy by Anchorage area electric utilities to rely on natural gas as a fuel for power genera­ tion could result in their turbines running out of fuel early in the 21st Century.

However, there is another side to the paradox that presents a dilemma to Alaska. It is ~uite possible that Alaska's remote location may result in Alaskans receiving abundant gas supplies at appreciably less than the cost of alternative energy supplies. If so, then the use of gas as a fuel for power generation could result in, by far, the least costly power for Alaska. The questions Alaska must answer with regard to future power generation, therefore, are (1) will gas be available where it is needed, and (2) will it be available at a price that allows economic power generation? The reason natural gas may be an abundant, low-cost fuel in the future is similar to the reason ilil natural gas currently is an abundant, low-cost fuel in Alaska. Anchorage residents enjoy the lowest natural gas rates in the United States, and because natural gas is used to generate electricity, they enjoy some of the lowest electricity rates as well. The reason relates to circumstances that determine natural gas prices. B.3.2 Pricing of Natural Gas

Natural gas often is described as a II superior" fue 1 because it is c1ean-burni ng and does not require user storage. Traditionally, however, when price distortions due to regulation are stripped away, natural gas has never commanded as high a price at the wellhead as crude oil. This paradox results from the fact that natural gas has higher long~distance transportation costs as well as higher distribution costs than oil. Natural gas often is discovered during exploration for crude oil, and produced with crude oil. Until sufficient gas reserves are discovered to justify construction of transportation facilities to distant markets, gas produc­ tion often greatly exceeds local needs, and gas sells at distressed prices. When gas is trans­ ported a long distance to markets, net-back prices from the point of competition generally cause gas to sell for less than alternative fuels. Natural gas prices normally are determined by one of the following methods: 1. The marginal cost of production from previously discovered reserves. This condition prevails whenever there is a large surplus, and producers compete to sell their gas in a limited market. 2. The marginal cost to discover and develop new reserves (when no surplus exists). 3. A net-back price from a distant marginal point of competition (where gas competes with other fuels). This condition normally prevails when local supply greatly exceeds local needs and the gas is shipped to distant markets. The first condition results in the lowest price. The second condition may allow (require) a higher price than the third. However, when large volumes of gas are transported long distance from a prodllci ng area, net-back prieing may either set a 1imi t on the price of gas or pull it higher. Gas reserves discovered in the Cook Inlet area were large compared with local needs, but were not sufficient to justify construction of a pipeline to distant markets. As a consequence of the large gas surplus, Anchorage atea electric and gas utilities have been able to purchase gas on long-term contracts at low cost in a buyer's market. In an attempt to obtain a higher price for their gas, Cook Inlet producers constructed two export facilities. One facility liquefies the gas and shi ps it to Japan as 1iquefied natural gas (LNG); the other faeil i ty converts gas to urea and ships the urea to the U.S. West Coast and foreign markets. Currently these two facili­ ties consume about two-thirds of the gas produced from Cook Inlet gas fields, excluding field 8-5

use and losses. However, reserves are still large compared with local needs, and producers have not yet obtained the market power to substantially raise grices. Present gas contracts in Alaska were negotiated in a buyer's market. Future gas contracts will more likely be negotiated under less ideal conditions from the buyer's point of view, although it is possible that the reserve/production (R/P) ratio in the Cook Inlet area may, as a result of new discoveries, remain high enough to keep prices low. ",. 8.3:3 Future Price of Natural Gas There are four possible scenarios of events that could result in somewhat different gas prices in the Railbelt area. These are:

1. Completion of the Alaskan Natural Gas Transportation System (ANGTS) as currently proposed. (This would make natural gas available in the northern Railbelt area.) 2. Completion of a gas pipeline to the Gulf of Alaska and construction of LNG facilities for shipment to Japan or the U.S. West Coast. 3. North Slope gas not available to the Railbelt area but facilities are constructed to export additional volumes of Cook Inlet gas. 4. North Slope gas not available to the Railbelt and no additional facilities are con­ structed to export additional Cook Inlet gas. Under the first two scenarios, the adequacy of supply is not a factor, and price is the only consideration relative to whether or not gas should be used for power generation. Under the last two scenarios, both price and adequacy of supply are considerations. 8.3.3.1 Completion "of the ANGTS If the ANGTS is completed, North Slope gas will compete with residual fuel oil for industrial markets in the northern United States. The cost of gas in the Fairbanks area will be the net­ back price at the marginal point of competition, likely in the Chicago area. It should also be equal to the North Slope wellhead price (determined on a net-back basis) plus the cost of trans­ portation to the Fairbanks area. Under present market conditions and projected costs of the ANGTS, the net-back price to Fairbanks would be negative--which is why plans for the pipeline have been "temporarily" delayed. If projected transportation costs do not decline significantly, or if oil prices do not rise appreciab'ly, then the "temporary" delay could extend for several decades. Assuming that the North Slope producers will not agree to sell their gas until the net-back price is positive, present transportation cost projections require market prices of approxi­ mately $10 per thousand cubic feet (Mcf) (in 1983 dollars) to ensure marketability of the gas and construction of the pipeline. In such a case the net-back price at Fairbanks would likely be close to $5.00 per Mcf, or higher. However, the Incentive Rate-of-Return (IROR) regulation adopted by FERC provided a strong incentive for the Applicant to inflate the cost estimates. It is quite possible that under the changed market circumstances the Applicant and producers will now discover methods to reduce the cost of construction. Tussing et al. (1983) discuss such a possibility. In addition to the possibility of building a pipeline at less cost than currently projected, the cost per Mcf transported could also be reduced if additional reserves are dis­ covered on the North Slope and the size of the pipeline is increased to handle a large volume. Any projection of net-back prices in Fairbanks following completion of the ANGTS at this point in time is speculative. However, considering the possibilities for reducing costs, an initial net-back price as low as $3.00 per Mcf or less seems conceivab"le. Although high compared with current gas prices paid by Railbelt utilities, such a price would be appreciably less than alternative fuel prices, and could allow electric power generation at considerably less cost than is likely to be supplied by any other means. Even if cost reductions are not possible, and the net-back price would be as high as indicated by current price projections, power could still be generated at less cost than is likely to be supplied by any other potential source of power. If one assumes that the ANGTS (or an equivalent pipeline) will be constructed, when is it likely to be completed? If cost reductions are not possible, then the projected pipeline is unlikely to be completed prior to gas prices rising to approximately $10 per Mcf in the Midwest indus­ trial market (expressed in 1983 dollars). This is unlikely to happen before oil prices rise to approximately $60 per barrel (in 1983 dollars), which could be a long time in the future. If substantial pipeline cost reductions are possible, and additional reserve discoveries allow transportation of greater volumes to achieve economies of scale, then a market price of $6 to $7 per Mcf might be sufficient to provide an economic incentive to construct a pipeline. This would require oil prices of $35 to $40 per barrel. If FERC oil price projections are correct, it could be well into the 21st Century before such an oil price is reached. Although oil prices 8-6

could rise above those projected by FERC, there appears sufficient doubt concerning future oil prices to cast considerable uncertainty on the potential availability of North Slope gas as a potential fuel for power generation in the Railbelt. 8.3.3.2 Completion of Gas Pipeline to Alaskan Gulf and Construction of LNG Export Facilities Should the gas pipeline be completed, gas would become available in essentially unlimited quanti­ ties in both the Fairbanks and Anchorage areas. The net-back price in Fairbanks would be the same as if the ANGTS were constructed, since it would still be equal to the net-back wellhead price plus the transportation cost to Fairbanks. The cost in the Anchorage area could be as low as $4 per Mcf or less, although this, too, is speculative. The principal difference from the ANGTS case is that liquefaction and transportation costs to Japan could be less than transporta­ tion costs to Chicago, which could allow an LNG project to become economic at a lower world oil price, and hence sooner, than for the ANGTS case. However, considering the problems LNG projects have had recently, and the risk that would be involved with a project of the magnitude necessary to market the large volumes of North Slope gas within a period of time that the sponsors would consider to be a reasonable market life, the prospects for completion of a pipeline to transport North Slope °gas to the Alaskan Gulf may be even more remote thanothe prospects for the ANGTS. 8.3.3.3 Construction of Facilities to Export Additional Volumes of Cook Inlet Gas A plan to export Cook Inlet gas to the U.S. West Coast has been actively considered and is still pending. Two California utilities and their subsidiaries, PacAlaska LNG and PacIndonesia LNG, filed applications with the Federal Power Commission (now FERC) in 1974. The utilities have defended their application against challenges on siting, environmental, economic, and safety issues, and the application is still pending before the FERC (Docket 75-140). However, the Indones i an reserves ori gi na lly dedi cated to the project have been re1i nqui shed and recent LNG sales from Indonesia have been at prices that could make export to the United States uneconomic. In addition, the option period for the 950 billion cubic feet (8cf) of Cook Inlet gas dedicated :!:lli to the contract has expired and the producers can now sell the gas to other bidders should they so desire (in fact, some gas appears to have been sold recently). Further, under the presiding Administrative Law Judge's initial decision (August 13,1979, Docket CP75-140), Phase 1 of the project cannot be authorized until 1.6 Tcf of proven reserves are dedicated to it. Phase 2 (contemplated to start a year after Phase 1) can be authorized only when another 1.0 Tcf are dedicated to the project. The 2.6 Tcf required represent only 78% of the project's require­ ments. Thus, the total requirements for the project are approximately equal to the present proven reserves in the Cook Inlet area. Currently, the necessary reserves are not dedicated to the project, declining oil prices are inhibiting sale of the gas in California, and there is still strong opposition to the project there. As a consequence, the short-term prospects of initiating the project certainly are not encouraging. However, conditions could change, and the PacAlaska project (or a similar project) could provide effective competition to electric utilities in bidding for additional volumes of Cook Inlet gas. In any such bidding, the consortium wishing to export the gas would have a di sti nct advantage, since they coul d offer contracts to the producers for much 1arger gas vol urnes. If large additional reserves are discovered in the Cook Inlet and export facilities are author­ ized, the likely net-back price in Alaska theoretically could be quite low. A producer might consider even a wellhead price of less than $1 per Mcf preferable to leaving the gas shut in. However, at present the necessary reserves are not available, and prices will have to rise above current levels to ensure exploration. 8.3.3.4 No Additional Facilities for Export of Cook Inlet Gas The condition of no additional facilities being built for export of Cook Inlet gas is likely if no new gas reserves are discovered, or if additional reserves are not discovered at a suffi­ ciently rapid rate to justify new export facilities, or if gas and oil prices are not sufficient to justify liquefaction of the gas and its transport as LNG. If oil prices fall, it may be close to a decade before the economics of LNG export begin to look favorable again. If so, the electric util ities may be_ successful in obtaining contracts for gas previously dedicated to PacAlaska.-However, they will have to compete with the existing LNG and urea plants in bidding for °the gas. Such competition for uncommitted gas should cause gas prices to rise, 1ikely resulting ln additional exploration. What would happen then would depend on the results of the exploration. If very large volumes were discovered, there would likely be sufficient gas to supply both the local market and an export market. In such a case, adequate quantities of gas should be available for electric Qower generation at relatively low net-back prices. If suffi­ cient gas were discovered to justi1y export, but not enough for both export and the local market, the producers 1i kely woul d opt for the 1arger volumes of the export market. In thi s case, gas might not be available for power generation. If insufficient gas is discovered to justify export facilities, there could still be sufficient gas to supply local needs well into the 21st Century. In such an event, the price of the gas would depend on the magnitude of reserves B-7

relative to consumption. With reserve/production (R/P) ratios great~r than about 15/1, gas prices would be low (relative to oil prices). If R/P ratios fell below about 15/1, prices would start rising toward equivalent oil prices. This in turn w~ld stimulate additional exploration, which, depending on results, could cause gas prices to decline. B.3.3.5 Future Gas Prices From the above discussion, it should be apparent that predicting future gas prices in the Rail­ belt is even more difficult and uncertain than predicting future 041 prices. It seems likely that gas prices in the Railbelt will continue to be less than oil prices. What is not certain is how much less, and if sufficient gas will be available for extended use in electric power generation. On the other hand, there seems to be an excellent possibil ity that the necessary volumes of gas wi 11 be available, and at a price sufficiently low to be the least expensive fuel for electric power generation. This, of course, is Alaska's current dilemma. Opting to use gas for power generation could be expensive if sufficient volumes of gas do not become available when needed. Conversely, choosing any other alternative could result in much higher pOwer costs than necessary, should gas be available. FERC's gas price projections are based on an assumption that sufficient volumes of gas will be discovered in~he Cook Inlet to meet the future power requirements of the lower-Railbelt area, and that the electric utilities will be able to obtain several contracts for such gas. The price projections are higher than net-back prices should be for decades, but eventually are projected to be somewhat lower. While the Staff considers its gas price projection to be reason­ able, and sufficient to ensure additional exploration, there is considerable uncertainty in both the underlying assumption of gas availability and the gas price projections. B.4 COAL Because the only significant market for coal within the Railbelt is as a boiler fuel for produc­ tion of electricity, it does not compete with electricity as an end-use energy source. Further­ more, unlike petroleum fuels and natural gas, coal as an energy source is not linked as directly to the price of crude oil. The reason that this has been and will likely continue to be the case is that coal is not a close substitute for oil. The major uses to which coal is likely to be put are the conventional ones--as a boiler fuel for producing industrial process heat and for powering steam turbines for generating electricity by the utility industry. It is the latter use that is the internal market for coal within the Railbelt. The export market for the Rail­ belt's coal will likely entail both uses for this resource. The developing export market in the near term is, however, as a fuel used in generating electric power. Within the Railbelt, coal will compete with other sources of electric power generation--residual fuel oil, distillate fuel, natural gas, and hydroelectricity. Even here, however, coal is not a close substitute in certain applications. Coal can only be used in plants using steam as the prime mover. Thus, it is not as well suited to providing peaking or load-following generation. Coal transportation and fuel-handling facilities typically require significant investments, and there are emissions problems with combustion of coal that dictate constraints on its use in electric generation. The minimum scale coal-fired generating plants foreseen for the Railbelt area are on the order of 200 megawatts (MW). Plants this size constitute significant increments to the stock of total generating capacity in the region and would require substantial lead times to construct. The decision by a utility in the Railbelt to invest in such capacity is thus a major one, and no coal plants are currently under construction. Therefore, internal Railbelt coal consumption faces an upper bound during the present decade, as determihed by the fuel requirements of . existing capacity. The export market, on the other hand, does not suffer from the same constraints. First, there are industrial concerns, as well as utilities, that are potential customers, Second, much of the current interest by utilities stems from their decisions to convert existing capacity from alternative fuels to coal-fired generation. The lead times for such conversions are not as great as those for new plant construction. It is this fuel conversion activity that is indicative of the manner in which coal markets are indirectly tied to the price of world oil. The industrialized nations of the world saw the costs of making coal a substitute for other fuels (particularly petroleum·fuels) suddenly fall relative to the costs of continued dependence on those other fuels. This change was, of course, due to the escalation in the price of world oil, beginning in 1973. Added to the price escala­ tion, however, was the uncertainty of oil supplies for a number of major industrialized coun­ tries. Thus, initiatives were undertaken by some to diversify their reliance on alternative energy sources. This has had the impact of increasing the worldwide demand for coal, and prices have climbed as a result, although not nearly as much as prices for petroleum fuel and gas. This represents the major link between coal markets and the price of crude oil. If crude prices cl imb, then the economic potential for substitution of coal for both oil and gas wi 11 ...... B-8

continue to increase; the market for coal will expand, and there will be upward pressure on the price of coal. The converse is equally true and, perhaps, more likely. With respect to the export market for Alaskan Rai"'belt coal resources, the same economic forces are at work. The industrialized nations looking to substitute alternative fuels for their uncertain supplies of crude oil are the Pacific Rim countries of Japan and South Korea. Coal has been one of the fuels these nations have chosen to use as substitute, and currently they import coal from South Africa, Austral ia, Canada, China, and the contiguous United States. These other sources of supply, as well as potential supplies from mines being opened in eastern Siberia, have made for substantial competition in Alaska's attempts to expand its Pacific Rim exports. Should the market develop for Railbelt coal exports, then the export price that coal commands will constitute the real cost of consuming that fuel locally. The outlook for such expansion is mixed. Fi rst, the compet iti on among coal supp 1i ers to the Pacific Rim is sub stantia1 and wi 11 increase in the near future. Second, the motivating factor for the diversification away from petroleum and into coal, among other fuels, has diminished measurably during the last 18 months as the outlook for real escalation in world prices has moderated and the prospects for falling crude prices have become reality. Thus,the value of the coal available for electricity genera­ tion within the Railbelt is" likely to be the cost of extracting and transporting it to the generator. Given the vast suppl ies available to serve both the domestic as well as export markets, there is no persuasive reason to anticipate that the real costs of supplying the coal will escalate. B.5 PEAT Alaska contains permafrost-free peat deposits estimated at 27-107 million acres [109-433 billion square meters (m 2 )] that represent more than half the total U.S. peat reserves. Forty-seven 2 illfll million acres (190 billion m ) are located 5 feet (ft) [1.5 meters (m)] or less from the surface. Some 30 million acres (121 billion m2 ) show promise as an energy resource. A 1980 survey by the U.S. Department of Energy investigated large peat fields located in three separate locations within the Railbelt (the Matanuska-Susitna Valleys, Fairbanks, and the Kenai Peninsula) and concluded that they constituted a potentially valuable source of fuel, particularly for remote communities. According to the Division of Energy and Power Development of Alaska, peat for use in steam electric generation plants appears competitive with coal priced at $2.00 per million Btu; however, developmental and operational issues associated with prototype plants would have to be addressed before commercial plants could be contemplated. B.6 GEOTHERMAL ENERGY Several areas of Alaska have geothermal potential, particularly areas near or within the Rail­ belt. To date, however, only a fraction of that potential has actually been tapped-'-in the form of hot springs used for space heating and resort spas. Such springs are located at Manley Hot Springs, Chenea Hot Springs, and Tolovana. A number of geothermal sites are being investigated for their thermal energy and electric generation potential. Areas containing hot igneous systems, in or bordering on the Railbelt, include Mt. Drum, Mt. Wrangell, and Double Peak. In most cases, however, geothermal heating systems currently are not economically competitive with conventional heating alternatives. Drilli~g costs are extremely high, and the resource value of geothermal energy is critically dependent on the proximity to the end user. The heat distribu­ tion system for these wells can increase costs by a factor of five or six. According to the Division of Energy and Power Development, estimates of heat distribution piping average about $150/ft ($500/m), so eVen a small village of 50 residences, each about 150 ft (50 m) apart, would pay more than $1 million for the distribution system alone. B.7 TIDAL POWER Tidal energy is potentially available in Alaska, primarily in the Cook Inlet areas of the Rail­ belt, where the height of tidal variation and the volume of tidal flow are sufficient to make tidal power projects practical. Tidal energy can be converted into electricity by capturing both the potential energy associated with the height of tidal fluctuations and the kinetic energy associated with the flow of tidal water in and out of a contained area. If all the potential anQ~kinetic energy of Cook Inlet were captured and made available to users in the Railbelt area' of Alaska, it would provide electric power for the entire region well beyond the year 2050. A study prepared by Acres American identified 16 sites in the Cook Inlet area whose total energy capacity exceeded 186,000 gigawatt-hours (GWh) , with a total potential capacity of 73 GW. The Division of Energy and Power concluded, early in 1983, that development of commer­ cial tidal power is more than a deca~e away. B.8 SOLAR ENERGY Solar energy is not regarded as a potential source of power within the Rai"lbelt, either in the form of photovoltaic energy or solar heat. Despite the long hours of daylight that characterize 8-9

the summers in the Rai1belt, the periods of greatest energy need are during the winter, when solar energy production in Alaska would be negligible. To justify even the projected low invest­ ment costs in solar devices, it would be necessary for su~h equipment to make substantial contri­ butions to the supply of energy when energy requirements are greatest.

REFERENCES FOR APPENDIX B

Tussing, A.R. 1983. ARTA energy insights. Pacific Oil Insights.- (November). Tuss; ng, A. R., et a1. 1983. ARTA; ns; ghts. Natural Gas Ins;ghts. (Summer).

~..a.-. '" DRAFT ENVIRONMENTAL IMPACT STATEMENT SUSITNA HYDROELECTRIC PROJECT, FERC NO. 7114

.~

APPENDIX C ENERGY CONSERVATION

Prepared by Federal Energy Regulatory Commission Staff

C-1

_-'A!tw_'"""'". _ '"IIIII_~"'~"_,•• ' ,d1 C-3

'"

APPENDIX C. ENERGY CONSERVATION ,;" ~-;

.'~

C.l ENERGY CONSERVATION AND THE NATIONAL ENERGY ACT OF 1978 Provisions of the National Energy Conservation Act (NECA) of 1978 may have relevance to demand forecasts and other matters with which this environmental impact statement is concerned. The NECA provides for the following selected items: 1. Utility Conservation Program for Resi dences. A program requi ri ng utilities to offer to thei r residentia1 customers energy audi ts that woul d identify appropri ate energy cOnserva­ tion and solar energy measures and estimate their likely costs and savings. Utilities also will be required to offer to arrange for the installation and financing of any such measures. 2. Weatherization Grants for Low-Income Families. Extension through 1980 of the U.S. Department of Energy (DOE) weatherization grants program for insulating lower-income homes at an authorized level of $200 million in FY 1979 and 1980. 3. Solar Energy Loan Program. A $100 million program administered by the Department of Housing and Urban Development (HUD) that will provide a support for loans of up to $8,000 to home­ owners and builders for the purchase and installation of solar heating and cooling equipment in residential units. 4. Energy Conservation Load Programs. A $5 billion program of Federally supported home­ improvement loans for energy conservation measures; $3 bi 11 i on for support of reduced­ interest loans up to $2,500 for elderly or moderate income families, and $2 billion for general standby financing assistance. 5. Grant Program for Schools and Hospitals. Grants of $900 million over the next three years to improve the energy efficiency of schools and hospitals. 6. Energy Audits for Publ ic Buil di ngs. A two-year, $65 mi 11 ion program for energy audits in local public buildings and public care institutions. 7. Appliance Efficiency Standards. Energy efficiency standards for major home appliances, such as refrigerators and air conditioning units. 8. Grants and Standards. Grants and standards for energy conservation in Federally assisted housing. 9. Loans. Federally insured loans for conservation improvements in multifamily housing.

10. Sola~ Demonstration Program. $100 million for a solar demonstration program in Federal buildings. 11. Conservation. Conservation requirements for Federal buildings. 12. Solar Photovoltaic System. $98 million for solar photovoltaicsystems in Federal facili­ ties. 13. Objectives and Reports. Industrial 'recycling targets and reporting requirements. 14. -Labeling. Energy-efficient labeling of industrial equipment.

C.2 CONSERVATION OF OIL AND NATURAL GAS--THE POWERPLANT AND INDUSTRIAL FUEL USE ACT OF 1978 The Powerplant and Industrial Fuel Use Act (PPIFUA) of 1978 has as its principal objective the conservation of oil and natural gas supplies. The following provisions of the PPIFUA should effect substantial reductions in the nation's oil and natural gas consumption and should accel­ erate the conversion of oil-fired and gas-fired electric utility plants to coal-fired facilities: 1. Prohibition of New Oil- and Gas-Fired Boilers. Prohibition against use of oil or natural gas in new electric utility generation facilities or in new industrial boilers with a fuel heat input rate of 100 million Btu's per hour or greater, unless exemptions are granted by DOE. C-4

2. Restrictions on Existing Coal-Capable Large Boilers. DOE authority to require existing coal-capable facilities, individually or by categories, to use coal and to require non­ coal-capable units to use coal/oil mixtures. 3. Restrictions on Users of Natural Gas for Boiler Fuel. Limitation of natural gas use by existing utility power plants to the proportion of total fuel used during 1974-1976, and a requirement that there be no switches from oil to gas. There is also a requirement that natural gas use in such facilities cease by 1990 (with certain exceptions). 4. Pollution Control Loan Program. An $800 million loan program to assist utilities to raise necessary funds for pollution control. 5. Supplementary Authority. Supplemental authority to prohibit use of natural gas in small boilers"for space heating and in decorative outdoor lighting and to allocate coal in emer­ gencies. 6. Other Provi sions. Fundi ng of several programs to reduce negative impacts from increased coal production, energy impact assistance, and railroad rehabilitation.

C.3 THE PUBLIC UTILITY REGULATORY POLICIES ACT OF 1978--RATE DESIGN, LOAD MANAGEMENT, AND REDUCTION OF THE GROWTH RATES IN THE DEMAND FOR ELECTRIC POWER The Public Utility Regulatory Policies Act (PURPA) of 1978 is directed at reducing the growth rate in the demand for electric power, the reduction in the need for new generating capacity, and conservation of fuels in short supply. The PURPA provides for the following: 1. Rate Design Standards. Eleven voluntary standards on rate design and other utility prac­ tices for consideration by state regulatory authorities and non-regulated utilities-­ includingtime~of-day-rates, seasonal rates, cost of service pricing, interruptible rates, prohibiting of declining block rates, and lifeline rates. 2. Cons iderati on of Rate Des i gn Standards. A requi rement that state regul atory authorities and utilities consider each standard within prescribed periods and determine if the standards are appropriate for conservation, efficiency, and equity, as well as consistent with state 1aws. Voluntary gui de 1i nes. with respect to the standards may be prescribed. 3. Retail Pol icies for Natural Gas Activities. Consideration by gas util ities of two stan­ dards--i.e., service termination procedures and advertising expenditures. A DOE study of the best rate design for gas utilities is also required. 4. Cogeneration. FERC rules favoring industrial cogeneration facilities and requiring utili­ ties to buy or sell power from qual ified cogenerators at just and reasonable rates. 5. Wholesale Provisions. FERC authority to require interconnections of electric power trans­ mission facilities, to order utilities to provide transmission services between two noncon­ tiguous utilities, and to report anticipated power shortages; FERC review of automatic rate-adjustment clauses. 6. Aid to States and Consumer Representation. Funding to assist state implementation and consumer intervention in proceedings. 7. Small Hydroelectric Facilities. Loan program to aid development of small hydroelectric projects. 8. Significant Miscellaneous Provisions. Authorization funding for the National Regulatory Research Institute; establishment of three additional university coal research laboratories; rules for conversion from use of natural gas to use of less desirable heavy fuel oils; emergency conversion of utilities and other facilities during natural gas emergencies; natural gas transportation policy, and rules for treatment of conserved natural gas.

C.4 RATE D~?fGN AND LOAD MANAGEMENT--THE NARUC RESOLUTION NO. 9 STUDY A study of rate design and load management as potential expedients for the reduction of demand peaks and the associated need for additional peaking capacity in electric utility systems was initiated in 1974 by the National Association of Regulatory Commissioners (NARUC). Resolution No.9, Appendix A (1974),"" of this Commission called for "a study of the technology and cost of time-of-day metering and electronic methods of controlling peak-period usage of electricity, and also a study of the feasibility and cost of shifting various types of usage to off-peak periods.'1 It resulted in a detailed research plan that focused on shifting and con­ trolling loads in a way that would lessen the growth of peak demand. The ensuing research c-s emphasized the development of time-differentiated rates based on alternative costing method­ ologies and the evaluation of various direct load control techniques. In mid-1976, NARUC requested a continuation of the research. ' A Rate Design and Load Control Study was sponsored by the Edison Electric Institute (EEl), the American Public Power Association (APPA), the National Rural Electric Cooperative Association (NRECA), and the Electric Power Research Institute (EPRI). These sponsors and NAR~C encouraged representatives of their groups to participate in the study. ; The November 1977 report states that "--the research findings confirm a generally held but heretofore untested hypothesis that load management may yield benefits." The research findings indicate the desirability of load management techniques in some cases, as discussed below. First, for a small but diverse sample of companies, bulk power supply costs were found to vary markedly by time of day and season. The study established that time-differentiated rates, which reflect these costs much better than non-time-differentiated or seasonally differentiated rates, are administratively feasible. Second, the study found that customer use of electricity is responsive to time-differentiated rates, although the exact degree of c-hange is uncertain. Whether based on accounting or marginal costs, time-differentiated rates should, therefore, tend to reduce peak demand growth and the average bulk power s~pply cost. If electricity rates reflect marginal costs or if these rates diverge from marginal costs to recognize that the prices of other relevant commodities (e.g., fuels) are not based on marginal costs, economic logic suggests that the resulting use patterns should conserve society's scarce economic resources. It remains, however, a matter for further research,experimentation, and analysis to determine whether, for individual customers in individual systems, the sensitivity of consumption to price is high enough to yield total benefits commensurate with the total costs of time-differentiated rates. Third, the research established that direct loan controls have, in selected instances, benefits that exceed their costs. It should follow that individual systems should investigate load con­ trols as a cost-effective method for curbing peak demand growth and the consequent capacity expansion requirement. The final report on Phase II of the NARUC Study will supposedly emphasize critiques of proposed methodologies applicable to rate design and load management. Results of cost-benefit analyses to be included in the Phase II Report are expected to provide a better understanding of the potential impacts of rate design and load management on future load shapes. Recognized experts in the field of rate-making and load management were quick to respond to the Phase II Report on the NARUC Study and their comments clearly indicated that the study left many unanswered questions and raised many new ones. Some of these comments appear in the December 1, 1977, issue of "Electrical World," pp. 21-22.

l:T~~'HARY, ALASKA 11',rrEI{;IOR U.S. 1

'" DRAFT ENVIRONMENTAL IMPACT STATEMENT SUSITNA HYDROELECTRIC PROJECT, FERC NO. 7114 ~>'?-~

APPENDIX 0 345-kV TRANSMISSION LINE ELECTRICAL ENVIRONMENTAL EFFECTS

Prepared by Federal Energy Regulatory Commission Staff

0-1 D-3

""

APPENDIX D. 345-kV TRANSMISSION LINE ELECTRICAL ENVIRONMENTAL EFFECF~< ~

D.l INTRODUCTION Transmission lines of practical design create high electric field gradients at the conductor surface that cause ionization of the surrounding air layers when the field intensity exceeds the breakdown strength of this air. The resulting corona formation on the conductors (along with random gap discharges on other line hardware) gives rise to radio noise, audible noise, and generation of ozone (0 3 ) and oxides of nitrogen (NO). Corona formation is a function of line voltage, conductor radius, line geometry, conductorXsurface condition (roughness, adherence of foreign particles, etc.), relative air density, humidity, wind, and precipitation. Corona and its associated audible and radio noise levels increase substantially during periods of foul weather, espetially rain. Hence, it is neither practica1 nor economi ca 11y feas i b1e to des i gri extra-high-voltage (EHV) lines such that they will never be in corona, as is accomplished at lower voltages, although lines are commonly designed with sufficient conductor size or bundling to limit surface gradients, within the normal operating voltage range, below the critical level at which corona begins to sharply increase. Energized, load-carrying transmission lines also generate electric and magnetic fields that permeate the surrounding medium and induce voltages and currents in conducting objects in the vicinity, including persons and animals. The question of potential hazards of these fields from a biological and environmental standpoint has been given increasing attention in recent years, particularly with regard to lines designed for operation in the EHV range [345 to 1,000 kilovolt (kV)] and for future lines being considered for operation in the ultra-high-voltage (UHV) range (above 1,000 kV). In assessing the environmental impact of the expected levels of these electrical effects for the Susitna project 345-kV transmission lines, due recognition should be made of the fact that such lines have been in existence in other parts of the United States for some 30 years now. These lines traverse sparsely settled rural regions as well as high-density populated areas. As a result of this development, the design of these lines with regard to known electrical effects and other environmental aspects has become well established. Furthermore, the 345-kV operating voltage lies near the lower threshold voltage level at which many of the electrical effects associated with higher voltage lines become of marginal significance. Nevertheless, the Alaska Power Authority (APA) had an analysis made to predict levels of elec­ trical effects from the proposed project transmission lines, calculated using methods developed at Project UHV (Electric Power Research Inst., 1975a). A survey also was made of existing radio and television broadcast signal strengths and ambient radio noise levels along the Anchorage­ Fairbanks transmission corridor* for use in evaluating the influence of some of these line­ generated electrical effects. In addition, a survey was made of sensitive communication facility locations in the vicinity of the corridor, such as microwave installations and air navigational radio beacons. Recommended minimum separation distances of those facilities from the lines were developed, based on existing guidelines and criteria. This study was performed by APAls con­ sultant, Commonwealth Associates, Inc. (CAl). The results are presented in the IIElectrical Environmental Effects Report R-2394 11 (Commonwealth Assoc., 1982). The presently planned routes and number of circuits ultimately to be installed as part of the Susitna project 345-kV transmission system are indicated in Figure D-1. The calculations used to develop the predictions of electrical effects discussed in Report R-2394 were based on three single-circuit 345-kV transmission lines on a common 400-foot (ft) [120-meter (m)] right-of-way (ROW), as shown in Figure D-2, operating at a voltage of 362.5 kV. This would be typical of the structure placement for the Knik Arm-Gold Creek section of the Anchorage-Fairbanks transmission corridor. Electrical effects generated by this particular transmission link should be representa­ tive of the entire 345-kV transmission configuration ultimately to be installed as part of the Susitna project, and the Report calculations should be conservative due to ~he multiple-circuit ROW occupancy represented and the upper limit of the normal operating voltage range, 362.5 kV (5% above nominal 345-kV level). Both of these factors tend toward increased intensity of such effects as audible and radio noise, and ozone production.

*Hereinafter referred to as the IIAnchorage-Fairbanks corridor,1I or simply IIcorridorli where the meaning is clear from the context. D-4

, /_. ~~~.., rUKON_,..P ...r--....."'-.•. ····------~~-_____,""" ....-___l_---~~~~~>O,L------~-I-_: .../ -- / /

LEGEND ooo(!>- PRIIlARY ""'VED UNDIVIDED HIGHWAY --- SECDNIWlY ""'vED UNDIVIDED HI,GHWAY ------SECON!WN GRAVEL HIGHWAY, - RAILROAD _ ...- RIVER = Eiii PRD~ECT TRANSIiISSION LINIS ~ PlIOPOIED DAII SITE 6. TIIANSIiISSIDII SUUTATIDN

O~~~2~Oiiiiiiiiiii;i40 MILES SCALE I:

'\. ...",~ . .~ ,,~ '\.."".;. '!'!.~ ..· '::~\.. ~r '-.I ". f/ WILLOW !r ,SUBSTATION \.

---

Figure D-1. Susitna Project 345-kV Transmission System (ultimate configuration). 0-5

'"

1-0/&" DIA. ENS.

,,' ,,·AVG. TOWER 1 \ WI '~;1 SPACING =1400 FT. 1 62' 62' 98.5' 85' 74' GUYED STEEL X-FRAME

EST. FOOT ING RESISTANCE =50.\1 ,. 105'------'- ----105' -I N.T.S, I­ '200'-,-,-,------200"------1.1 I- 400' _I EDGE OF R.O,W. R.O.W. EDGE OF R. O. W.

Notes: Dimensions indicated are in feet (') and inches ("). 1 ft. = 0.305 m. Average Conductor Sag = 37 ft. **Minimum Ground Clearance at RR Crossing =38 ft. The 30-ft minimum phase-to-ground clearance is recommended in the APA Anchorage-Fairbanks Transmission Intertie basic design criteria, dated September 1981, pg. 6 (Prepared by Commonwealth Assoc., Inc.). This will be the design clearance fit 120°F. For railroad crossings, this clearance will be increased to 38 ft. The 44.7-ft minimum average height is equal to the minimum conductor-to-ground clearance (30-ft) plus one-third the conductor sag. (Transmission Line Reference Book, 345 kV and above,2nd. ed., 1982, Electric Power Research Inst., pg. 219.) This is the height commonly chosen on which to base line electrical performance design curves, stich as radio noise generation.· The 63.3-ft minimum shield wire height is based on a shield wire sag of 80% of the shown maximum conductor sag (44 ft) (the APA Shielding and Shield Wire Coordination Study for the Anchorage-Fairbanks Transmission Intertie, January 1982, pg. 4.) The subconductor and bundle size for the phase conductors and the shield wire size to be used are given in the APA Environmenta·l Assessment Report, R-2422, dated March 1982, pp. 18, 27. ' All other information taken from Application, dated February 1983, Exhibit F (Plates F80, F8l), and from the APA structure study for the Anchorage-Fairbanks Transmission Intertie, dated September 1981.

Figure 0-2. Typical Tangent or Light-Angle Structure Placement Along Knik Arm-Gold Creek Section of Anchorage-Fairbanks 345-kV Transmission Corridor. 0-6

The following sections contain a brief discussion of the nature of the environmental effects produced by transmission line operation, along with an assessment of the environmental sig­ nificance of these effects, particularly with regard to the relevant parameters of electrical effects calculated for the Susitna project transmission lines in Report R-2394. Guidelines used in making this assessment consisted of material contained in the referenced report, as well as reference information and data on this subject developed by the Electric Power Research Institute (EPRI) and others. The following topics are covered: Ozone and other air pollutants generated by transmission line corona, Audible noise generated by transmission line corona, Radio noise generated by transmission line corona, and minimum separation distances between the 345-kV lines and existing communication towers, Electrostatic and electromagnetic field strengths set up by transmission line operation, and related field effects, and Line clearances and electrical safety. As a result of this review, the Staff concurs with the electrical environmental effects analysis and conclusions reached by APA/CAI; i.e., no adverse environmental consequences of a permanent and irremediable nature should result that could be attributed to the operational performance of the 345-kV transmission 1ines that would be constructed as part of the SusHna project. Specifically, the following qualitative assessment is made: 1. No environmentally hazardous levels of corona-generated ozone or oxides of nitrogen should result from operation of the lines; in fact, the resulting increment to ambient levels due to line operation would likely not even be measurable. 2. Audible noise generated by corona formation on the lines would not be objectionable and would not contribute significantly to ambient noise levels. 3. Corona-generated radio noise would not likely interfere with AM radio broadcast reception at distances greater than 1,000 ft (300 m) from the centerl ine of the right-of-way even under worst-case weather conditions for noise generation, viz., during rain. No inter­ ference at all would be expected for FM radio reception due to its inherent noise rejection capability. Television reception should be unaffected at locations where television recep­ tion is presently good. Problems would be expected to arise only rarely, if at all, and mitigative measures could generally be employed to alleviate them on a case-by-case basis (such as by relocating receiving antennas). The routing of the lines would be adjusted as necessary to allow for industry-recommended separation distances from sensitive microwave and other types of communication facil ities to avoid potential interference problems. 4. Results of studies on possible biological harm from exposure to electric and magnetic fields are inconclusive at best, and no general acceptance of such a correlation seems to exist among the scientific community (Bridges, 1975; Sheppard and Eisenbud, 1977; Riog, 1979; Electric Power Research Inst., 1982; Mahmoud and Zimmerman, 1982). Inasmuch as the proposed line design conforms to generally accepted and long-established design practice for 345-kV transmission lines, the same normal levels of field intensity at ground level would result from these lines as for all the other numerous existing 1ines in this class. It is therefore concluded that no reasonable basis for concern exists On this account. Likewise, no shock hazards from induced potentials due to these fields would be expected. 5. The 30-ft (9-m) minimum phase-to-ground clearances are more than sufficient to satisfy the present requirements of the National Electric Safety Code, including the 5 milliampere induced current limit on large vehicles short-circuited to ground under the lines. Again, this conforms to present and long-established design practice for lines in the 345-kV class. The foregoing conclusions apply for the Susitna lines operated within the normal ±5% limits of their nominal design voltage level, 345 kV. Initially, the first transmission link, currently bei ng construc:t~d along the Anchorage-Fai rbanks corridor, wi 11 be operated at 138 kV, at whi ch voltage the levels of the foregoing electrical effects should be entirely negligible. D.2 OZONE PRODUCTION .... Operation of EHV transmission lines of practical design causes the formation of corona around the line conductors. Corona consists of ionized air particles in the adjacent air layers. Corona is formed when the electric field gradient at the conductor surface exceeds the breakdown strength of the adjacent air. Ozone and oxides of nitrogen, by-products of this ionization, add to the ambient atmospheric concentrations of these oxidants. The latter are created by natural 0-7 processes, primarily by the action of ultraviolet light from the sun on upper atmospheric air layers and on automotive and industrial emissions near the earth1s surface. They also are generated by lightning discharges. Ozone is the most impo~tant of these products from an envi­ ronmental standpoint because it comprises 80% to 90% of the atmospheric oxidants (Electric Power Research Inst., 1975b), whether produced from natural or man-made processes. In concentrated form, it is a powerful oxidizing agent with high chemical reactivity. Ambient levels of ozone generally vary from about 0.01 to 0.03 parts per million {ppm) in rural areas, although concentrations up to 0.10 ppm have been measured in-some rural areas (Bonneville Power Admin., 1977). In urban areas, concentrations of 0.10 ppm can be expected, and in some cities, such as Los Angeles, concentrations as high as 0.5 ppm have been measured, apparently due to the high levels of auto emissions and industrial combustion (Bonneville Power Admin., 1977). However, a number of investigations carried out over the past decade, including both field and laboratory test programs, have shown that practically no measurable incremental con­ tribution to ground-level concentration of ozone and oxides of nitrogen result, under any weather condition, in the immediate vicinity of transmission lines designed for operation not"only at 345 kV, but also at voltages through 765 kV (Frydman et al., 1972; Scherer et al., 1972; Frydman and Shik, 1973; Fern and Brabets, 1974; Sebo et al., 1975). These results are based on state­ of-the-art instrumentation accuracy of 0.001-0.002 ppm. In interpreting these results,it should be kept in mind that ground-level concentrations of ozone and other oxidants are a func­ tion of not only the generation rate (whether by natural or man-induced processes), but also the rate of decay and diffusion in the atmosphere. Ozone, for example, is one of the most chemically reactive agents known. It is therefore very unstable and is readily consumable by plants, animals, nitric oxide in the atmosphere, and by other substances, although at different rates. Under normal atmospheric conditions, ozone has a characteristic half-life of ~ to 1 hr (Frydman and Shik, 1973). Its rates of generation and decay are functions of temperature, humidity, initial concentration level, sunlight, rainfall, and other factors. As a result, ambient ozone concentrations are subject to large daily variations of as much as 0.08 ppm, with the highest concentrations occurring during daylight hours and the lowest at night (Electric Power Research Inst., 1982). The ground-l eve 1 concentration of ozone near transmi ssi on 1i nes is also a func­ tion of wind speed, direction of wind relative to the line direction, and line height. Trans­ verse winds result in greater diffusion, and the ground-level concentration diminishes with increases in wind speed and line height, It follows that ground-level concentrations of ozone and other oxidants generated by transmission line operation are not cumulative over time. Ozone production increases with an increase in corona level, which for a given line design is a function of the operating voltage, surface conditions of the conductor, and atmospheric condi­ tions. As has been indicated, as long as a line is operated within its design voltage limits, the maximum corona generation, and therefore ozone production, is fairly predictable. But a sufficiently high overvoltage, say, on a long line under light loading conditions, could exceed the critical level at which corona generation begins to increase much more sharply with voltage.* Under these conditions, the ozone generated by the line would increase correspondingly. However, sufficient reactive power sources are normally provided for transmission system operation to maintain the voltages within fairly close tol~rances (about ±5%), primarily for equipment protec­ tion and other operating reasons, but which also attends to the matter of ozone production. However, with the miniscule production of ozone at normal operating voltage, it is unlikely that any degree of sustainable steady-state operating voltage would result in an environmentally objectional production of ozone from_t/~e line. Corona loss in foul weather, particularly rain, is typically at least one order of magnitude greater than the corresponding fair weather loss on a given line, resulting in increased ozone production. However, field tests showed that the actual ozone concentration at ground level under rain conditions decreased, indicating that the accelerated decomposition of the ozone caused by the increased moisture more than compensated for the increased production rate due to the increased corona level (Fern and Brabets, 1974). The toxicological effect of ozone on humans has b~en investigated, and a maximum allowable concentration of 0.1 ppm for eight hours continuous exposure per day, five days per week has been establ ished by the American Conference of Government Industrial Hygienists (Fern and Brabets, 1974). However, the EPA air quality standards allow a maximum concentration of ozone of 0.12 ppm, not to be exceeded more than one day per year (U.S. Environmental Protection Agency, National Primary and Secondary Ambient Air Quality Standards for Ozone, 40 CFR 50.9). Reports of studies performed by EPA (Bonneville Power Admin., 1977) and others indicate that human respiratory tract irritation occurs at ozone levels between 0.5 and 0.7 ppm. Small laboratory animals developed chronic bronchitis when exposed to doses of 1 ppm for one year. Insofar as its effect on vegetation and ground foliage is concerned, studies on selected species known to be sensitive to such effects showed that a concentration of 0.07 ppm lasting four hours was

*Overvoltages from switching surges and lightning strikes could also cause this, although they are much too transitory in nature to be of significance with regard to ozone production. 0-8

requi red to damage eastern white pi ne, and concentrations of 0.10-0.12 ppm 1asti ng two hours were required to damage sensitive varieties of alfalfa, spinach, clover, oats, radish, corn, and beans (Bonneville Power Admin., 1977). The levels cited are well above the minute incremental concentrations caused by EHV transmission line operation in their immediate vicinity, but are not always above the ambient levels of ozone in some areas, as previously indicated. In summary, based on results of recent investigations, it appears that no significant levels of ozone would be produced by the project 345-kV transmission lines, even for as many as three 345-kV lines on the same right-of-way. Existing oxidant limits imposed by Federal and state agencies should create no difficulties for these or other EHV transmission lines, although existing ambient levels are close to or exceed the limits being set in some locations. 0.3 AUDIBLE NOISE High-voltage transmission lines generate audible noise as a result of corona formation along the 1i ne conductors. The noi se produced cons i sts of two pri ncipa1 components: (1) a broadband noise created by the random pulse discharges in the air at the surface of the conductor, and (2) a low-frequency pure-tone noise (hum) predominantly at a frequency of 120 Hz and created by the alternate attraction and repulsion of positive and negative ions (generated by the corona) under the action of the alternating electric field. The main contributor to annoyance is from random (broadband) noise in the 1-8 kHz range (Electric Power Research In~t., 1982). . Audible noise, like the corona that produces it, depends significantly on prevailing weather conditions for a given line geometry and operating voltage. A person standing under or near a line built to acceptable design standards for foul weather corona probably would not be aware of any audible noise in fair weather unless he were listening for it. Any appreciable air turbu­ lence probably would mask this effect. The fair weather noise level is, therefore, generally of no concern (Electric Power Research Inst., 1982), typically lying considerably below that for foul weather, particularly for rain when audible noise generation is highest due to the localized high electric field gradients formed at the water droplets. However, in assessing the potential disturbing effect of transmission-line-generated audible noise, account must be taken not ohly of the generation rate, but also of the rate of attenuation of the noise with distance from the line, the absorption of sound energy by the surrounding air, the masking effect of other envi­ ronmental noise sources, and the relative level of pUblic activity and degree of 'exposure to the noise under weather conditions that produce high noise levels (rain and intense fog). The sound , pressure level of the noise varies inversely as the square of the lateral distance from the line due to the divergence of the sound pressure waves (Electric Power Research Inst., 1982). However, the actual attenuation is somewhat greater than this due to atmospheric absorption, which increases with frequency and is also a function of air temperature and relative humidity. During rain, background noise from wind, thunder, and the rain itself, combined with reduced outdoor public activity and the reduced possibility of direct public exposure to the noise through open wi ndows near the 1i ne, all woul d tend to reduce any di sturbi ng effed of the increased noise generaiion rate: However, at other times, such as during periods of snow, heavy fog, or immediately after a rain, the outdoor environment often becomes quiet, making the increased audible noise generated by the water drops on the line relatively more evident, even though the noise generation rate is somewhat less than during heavy rain. This "wet conductor" condition closely corresponds to the mean noise level in rain (Electric Power Research Inst., 1982). For these reasons, it is often us'ed for assessi ng the audi bl e floi se performance of AC transmission lines, at least in the absence of more comprehensive statistics on audible noise. The range of sound pressure levels that the human ear can detect and assimilate is on the order of 1,000,000/1. For this reason, these levels are customarily expressed on a logarithmic scale by relating the measured sound pressure from a source, in micropascals (~Pa), to a reference sound pressure of 20 ~Pa (where 1 ~Pa =1 ~N/m2 =10- 5 ~bar) and taking 20 times the logarithm to the base 10 of this ratio, i.e., Sound pressure level (SPL) = 20 10g10 SP~o(i~ ~Pa) dB. -(0-1)' in decibels (dB) ~ a Some common noise levels are given in Table 0-1. The range of fregQencies audible to the human ear is about 15-20,000 Hz. However, human hearing is more sensitive to the range of frequencies in which most speech information is carried, or from about 500-4000 Hz, and falls off fairly sharply beyond these limits (Electric Power Research Inst.,1982). In measuring broadband noise, it is therefore customary to apply weighting to the different frequencies contained in the monitored noise in accordance with this characteristic such that the overall sound pressure level measured will relate as closely as possible to the ear's perception of the sound. This is commonly referred to as the A-weighting network in standardized sound level instruments, and the corresponding meter scale reading is referred to in units of dB(A), as in Table 0-1. Standardized B, C, and 0 weighting networks also are avail­ able that have different frequency response. characteristics suitable for measuring impulsive and other types of sound, some of which have been advocated for measuring transmission line noise. However, the A-weighted network is by far the most widely used noise-rating scale. 0-9

Table O-l. Noise Levels of Typical Noise Sources ,... Operatort1 Communityt 2 Noise Source [dB(A)] [dB(A)] Air conditioners 70-96 52-77 '.> Power lawn equipment 80-95 59-85 Chain saws 103-115 64-86 Automobiles 55-87 77-87 Snowmobiles 100-116 78:::'88 Motorcycles Less than 240 cc 90-105 70-90 Greater than 240 cc 95-115 78-95 Trucks 70-100 70-95 t 1 Operator: Noise levels measured at the position of the operator of the noise source. . t 2 Community: Noise levels measured at locations 50 ft (15 m) from the center line of the path of the source or 50 ft (15 m) from the source. Source: Anonymous (1974).

Inasmuch as many sounds, including transmission line noise, have sound pressure levels that are not constant with time, they cannot adequately be characterized by a single value of SPL. To deal with this problem, statistical correlation with time is often resorted to, using indices such as Ls , indicating the A-weighted sound level that is exceeded only 5% of the time; Lso for A-weighted sound levels exceeded 50% of the time, etc. (Electric Power Research Inst., 1982). (The Ls level corresponds to heavy rain generation, and the Lso corresponds to wet conductor generation.) If a noise is intermittent or fluctuating, an equivalent sound level, L ,is used, defined as the energy average (usually A-weighted) of a varying sound over a splc~fied period of time; i.e., a steady sound having the same level as the L would have the same sound energy as the fluctuating sound. This is a useful measure in conne~~ion with transmission line noise measurements; however, it does not account for the more annoying effect of noise at night. For this purpose, a modified L has been developed, designated as L for day-night level (U.S. Environmental Protection ~ency, 1974). It adds a 10 dB penal~9 for noise occurring between the hours of 2200 and 0700, and is calculated as: 2~ 10 Ldn = 10 10g10 [15 (10 Li ) + 9 (10 (Ln + 10)/10)] dB(A) , (0-2) where

Ld = Leq for daytime hours (0700-2200) and Ln = Leq for nighttime hours (2200-0700). Transmission line audible noise levels can be estimated for design purposes based on empirical formulas developed from laboratory and field measurements that correlate such factors as con­ ductor geometry, tower design, conductor surface gradient, and distance from line to the measur­ ing point to determine a value for sound pressure level in dB(A). The audible noise levels for the project 345-kV lines have been calculated using methods developed at Project UHV (Electric Power Research Inst., 1975a). The calculated noiSe levels at the edge of a right-of-way con­ taining three single-circuit 345-kV lines under heavy rain and wet conductor conditions, and at maximum operating voltage of 5% above normal, or 362.5 kV, are given in Table D-2. Estimated fair weather levels (inaudible) also are given. Using the wet-conductor levels as a measure of the line audible-noise performance, the formulation of Eq. 0-2 was used to calculate an equivalent day-night level, L ' at the edge of the right-of-way under the very conservation assumption that the conductorsnwillrl remain wet for 24 hours, i.e., L = L =44 dB(A). Allowing for the 10 dB nighttime penalty incorporated in Eq. 0-2, this resJilted"in a day-night average (L ) of 50.4 dB(A). These computer-calculated numbers, taken from Report R-2394, are in reasoni~ly close agreement and are conservative with respect to corresponding values of 41.7/48.1 dB(A) computed by the FERC Staff using manual methods based on design curves and formulas given by the Electric Power Research Institute (1982). 0-10

Table 0-2. Calculated Audible Noise levels for the Anchorage-Fairbanks Corridor with Three 345-kV Lines on a Common Right-of-Way Operating at 362.5 kV

L at Edge of Weather Conditions Right~8f-Way [dBCA)Jtl Heavy rain 54 Wet conductor 44 Fair weather Inaudiblet2 t 1 200 ft (61 m) from centerline. t 2 Estimated. Source: Commonwealth Associates (1982).

So far as is known, no existing noise limits specified by ordinance, regulation, or statute specifically refer to transmission lines. However, based on summaries of guidelines developed by the U.S. Environmental Protection Agency as given in Table 0-3, by the Committee on Hearing, Bioacoustics, and Biomechanics (CHABA) of the American Academy of Sciences as given in Table 0-4 and by the Bonneville Power Administration as given in Table 0-5 (all of these relate to the effects of noise and community reaction and annoyance), the day-night average of 50.4 dB(A) calculated for the 345-kV lines at the edge of the right-of-way should present no environmental problem. Although this value represents a day-night average,it is unlikely that peak noise levels would exceed the 52.5 dB(A) threshold level given in Table 0-5, except possibly for very limited periods in heavy rain when the lines were simultaneously operating at maximum voltage (362.5 kV). A comparison of the 50.4 dB(A) average with the noise levels given in Table 0-1, which represent levels to which people are normally exposed, suggests that the audible noise generation by the line should not contribute significantly to these levels. 0.4 RAOIO NOISE

ll ll IIRadio Noise (RN), sometimes referred to as lI electromagnetic interference (EMI), is a rather general term used to refer to any unwanted interference of an electromagnetic nature (such as radio static) with any signal or communication channels or devices throughout the radio frequency band of operation (3 kHz to 30,000 MHz). The pulsative corona discharges produced by energized high-voltage transmission line' conductors generate such disturbances by virtue of the steep rise and decay rates of the mi nute components of currentfeedi ng thi s corona and by _the rapid and random repetition rate of the corona pulse discharges along the line conductors and other line and substation hardware. Radio noise can also be generated by sparking at loose or broken line hardware parts. Theoretically, cotona-generated radio noise could cause interference with virtually any type of radio reception. However, in practice it has been found that the bands principally affected are the AM (amplitude-modulated) broadcast band (535 to 1,605 kHz) and the video signals of the low television bl70adcast band (Channels 2 to 6, 54 to 88 MHz). FM (frequency-modulated) radio signals (88-108 MHz), which are also generally used for the sound transmission for television, are virtually immune to the static-type interference generated by transmission line radio noise. This is because (1) the magnitude of the radio noise is generally quite small in the FM broad­ cast band, and (2) FM radio systems inherently reject this pulsative-type noise. For a given operating voltage, radio noise generation from a transmission line is principally a function of conductor geometry, conductor height above ground, phase spacing, and ground resistiv­ ity. Since it is a product of the line corona, it also depends on the condition of the conductor surface, increasing with roughness and contamination, and on weather conditions, becoming several orders of magilitude greater in rain than during fair weather. The radiated radio noise is broadband irf~gcharacter, with a decreasing frequency spectrum. For example, the magnitude typically decreases on the order of five to six times (around 15 dB) per decade of frequency, measured 400-500 ft (122-152 m) from the edge of the right-of-way, and becomes quite low in the frequency range above 10 MHz (Electric Power Research Inst., 1982). The radio noise level also attenuates with lateral distance from the line, typically at rates varying from 10 to 30 times (20-30 dB) per decade of distance, depending on the measuring frequency (Electric Power Research Inst., 1982). Interference caused by this noise to radio and television receivers in the vicinity therefore depends on their proximity to the line as well as on the signal strength of the desired o-n

'"

Table 0-3. Summary of Noise Levels Identified by USEPA as Requisite to Protect Public Health and Welfare with an Adequate Margin of Safety

Maximum Allowable Avoided Effect Level [dB(A)Jt1 Area

Outdoor activity interference L ~ 55 Outdoors in residential areas and farms, and annoyance dn and other outdoor areas where people spend widely varying amounts of time, and other places in which quiat is a basis of use

L ~ 61.4 Outdoor areas where people spend limited dn amounts of time, such as school yards, playgrounds

Indoor activity interference L ~ 45 Indoor residential areas and annoyance dn L ~ 51.4 Other indoor areas with human activities dn such as schools t 1 · Notes:

1. Ld is a weighted day-night average noise level with a 10 dB penalty added to the night­ tr~e equivalent noise level (2200-0700 hours).

2. Anchorage-Fairbanks corridor estimated Ldn at edge of right-of-way = 50.4 dB(A).

3. An indoor Lri of 45 dB(A) wi 11 permit speech communication in the home, whi 1e an outdoor Ldn not exci:!Qding 55 dB(A) will permit normal speech communication at approximately 10 ft (3 m). 4. Maintenance of L of 55 dB(A) will provide an indoor L of approximately 40 dB(A) with windows partly oSQn for ventilation. The nighttime por~~on of this L (indoor level) will be approximately 32 dB(A), which should, in most cases, protect ri~8ainst sleep interference. . Source: u.S. Environmental Protection Agency (1974).· 0-12

Table 0-4. Summary of Human Effects for Outdoor Day-Night Average Sound-Level of 55 dB(A)

Type of Effect Magnitude of Effect Speech Indoors No disturbance of speech: 100% sentence intelligibility (average) with 5 dB margin of safety Outdoors Slight disturbance of speech with 100% sentence intelligibility (average) at 0.35 m, or 99% at 1.0 m, or 95% at 3.5 m

Average community reaction None; 7 dB below level of significant ~complaints and threats of legal actionll and at least 16 dB below ~vigorous action~ (atti­ tudes and other non-acoustical factors may modify this effect) High annoyance Depending on attitude and other non-acoustical factors, approx­ imately 5% of the population will be highly annoyed Attitude toward area Noise essentially the least important of various factors Conversion: to convert meters to feet, multiply by 3.28. Source: National Academy of Sciences (1977).

Table 0-5. Audible Noise Complaint Guidelines Developed by Bonneville Power Administration

Audible Noise Level 100 ft Laterally from Line Centerline Probabil ity of Lso (wet conductor) [dB(A)] Receiving Complaints 0-52.5 Few or no complaints 52.5-58.5 Moderate or some complaints >58.5 High or numerous complaints Conversion: To convert feet to meters, multiply by 0.305. Source: Bonneville Power Administration (1977). D-13

incoming signal, its frequency, and the ambient radio noise level.* Depending on those factors, the magnitude of corona-generated noise can range anywhere from barely discernible levels to a point where reception is completely unintell igible. In lines of modern design, interference is likely to be most noticeable in AM receivers located close to or under a line, such as in an automobile passing under the line,and possibly in television receivers in proximity to the line in the far-fringe areas of reception. ,;, . " ; .., Radio noise is measured with standardized radio noise meters that, in essence, ate calibrated radio receivers that function as radio frequency voltmeters. They are capable of measuring radi 0 frequency noi se down to fractions of a mi crovoIt. The annoyance 1eve 1 of transmi ssion line radio noise on communication receivers is characterized in terms of a signal-to-noise ratio (SNR), defined as the ratio of the average signal level field strength, in microvolts per meter (~V/m), to the quasi-peak level of the noise, measured in the same units. The quasi-peak is a level intermediate between the peak and average noise levels. It is a standardized measure that accounts not only for the amplitude but also the repetition rate of the noise pulses,andthus more accurately represents the nuisance value of the radio noise field with respect to broadcast reception. Because of the wide range of possible signal and noise levels, a logarithmic scale is used to express these levels in decibels (dB and dBn' respectively), based on a 1 micrQvolt per meter (1 ~V/m) reference level, as s

Vs dB s = 20 log 10 -r dB~** (D-3) V n dB n = 20 log 10 -r dB~ (D-4) whence

SNR = dB - dB = 20 log 10 ~s dB~ (D-5) s n n

~V/m. where Vs and Vn are the signal and noise field strengths, respectively, in Estimates of expected radio noise levels generated by transmission lines can be calculated from empirical formulas and design curves that have been developed from data obtained in laboratory and field test investigations. In the United States, there are no'regulations on a local or Federal level that expressly limit the level of radio noise that a transmission line may produce, although Federal Communications Commission (FCC) rulings require, in general, that no device that radiates radio frequency energy shall endanger or seriously degrade the function of radio navigation services or radio communication service& (Federal Communications Commission, 1975). Technicafly, transmission lines fall within this category. Tolerability criteria for trans­ mission line-generated noise must therefore be based on subjective ratings of listeners and viewers exposed to radio, and television programs containing various amounts of injected or measured ambient noise. A preconstruction survey of signal strengths of AM radio broadcasts was made by the Alaska Power Authority (APA) through their consultant, ~ommonwealth Associates, Inc. (CAl), at 11 sites near the right-of-way of the 345-kV Anchorage-Fairbanks transmission corridor in the section between Willow and Healy (Commonwealth Assoc., 1982: Fig. A-I). A list of stations received at various sites is given in Table D-6, and the site locations are listed in Table D-7. All of the station signal levels were below the 40 dB~ level required by the FCC for primary service in the northern rural areas of Alaska, defined as those areas in which the ground wave is not subject to objec­ tionable interference or fading (Federal Communications Commission, 1968). The maximum signal strength of any of the stations monitored at the eleven sites was only 37 dB~ (KOFD, in Anchorage). Ambient radio noise levels were also measured at each of these sites at this time at selected frequencies in the AM broadcast band. The measurements were made during daytime under overcast or rain conditions. The resulting SNR varied from about -8 to +12.5 dB. While these readi-ngs were being taken, a subjective rating of each radio signal received was made by the team of observers on a scale ranging from "All (ll en tirely satisfactoryll) to IIE II (II speech unintelligiblell ). The results are tabulated in Table D-7. At two of the measuring sites, no AM broadcast signals at all were heard, and, at best, only IIC II (llfairly satisfactoryll) signal s were heard, with background noise plainly evident at three locations. Of the total of 32 signal receptions at the 11 monitoring sites (with a given station often received at more than one site), the 18 signa1s with the II Ell qual i ty correlated wi th an SNR range of -8.8 to -1. 2 dB~. The ni ne signals with a IID II quality correlated with an SNRrange of -4.4 to +3.0 dB~, and the remaining six signals with a IIC II quality correlated with an SNR range of +4.2 to 12.5 dB~. On this basis, an

*Ambient radio noise originates from atmospheric or other man-induced sources. A sufficiently high level of ambient noise could render the corona-generated noise component negligib'le by comparison. **dB~ signifies dB relative to 1 microvolt per meter. ~ 0-14

! Table 0-6. AM Radio Stations Received During Preconstruction Survey of Anchorage­ Fairbanks Transmission Corridor between Willow and Healy, July 1981

Frequency Station Powert 1 Antenna Station (kHz) Call Location (kW) Limitationt1 Classt2 550 KENI Anchorage 5 III 560 KVOK Kodiak 1 III 580 KYUK Bethel 5 III 590 KHAR Anchorage 5 III 650 KYAK Anchorage 50 DA-2 II 660 KFAR Fairbanks 10 II 700 KBYR Anchorage LS-l, N-. 5 II 750 KFQO Anchorage LS-50, N-I0 II 900 KFRB Fairbanks· 10 II 970 KIAK Fairbanks 5 III 1080 KANC Anchorage 10 II 1150 KABN Long Island 5 III (Big Lake) 1170 KJNP North Pole 50 DA-N II t 1 Key: DA-2--0irectional antenna, different patterns day and night OA-N--Directional antenna during night only LS--Local Sunset N--Night. t 2 Only Class II and Class III stations were received. Class II stations are licensed by the Federal Communications Commission (FCC) to operate on a clear channel and render primary service over wide areas. Class III stations are licensed by the FCC to operate on a regional channel and render primary service to large cities (municipalities) and surrounding areas. The primary service area is the area in which the radio signal is not subject to objec­ tionable interference or fading. Source: Commonwealth Associates (1982).

Table 0-7. Existing Quality of Reception for AM Radio Stationst1

Number of Radio Stations Judged to Have the Following Quality of 2 Site Audio Receptiont Number Location A BC 0 E

10 Wi 11 ow 3 3 20 Trapper Creek 2 2 3 30 Chase 1 4 40 Lane Creek 1 1 4 50A Curry 1 60 Cantwell 1 70 Carlo Creek 1 80 Oeneki Lake 1 3 90 McKinley Village 100 McKinley Park 110A Healy 1 1 .<~. ,::;Tota 1 0 0 6 9 18 t 1 Based on field measurements of radio station signal strengths July 9-15, 1981. t 2 A--entirely satisfac~ry; B--very good, background unobtrusive; C--fairly satisfactory, background plainly evident; O--background very evident, speech understandable with concentration; E--speech unintelligible. Source: Commonwealth Associates (1982). 0-15

SNR of +4.2 dB~ is about the minimum acceptable level,which was met by only six signal reception . points. Reception of the remaining 27 would be considered below par or unsatisfacto~ under this standard, even with presently existing ambient conditfbns and without the installation of the Intertie. An estimated lateral profile of radio noise levels along the corridor was calculated by APA/ CAl (Commonwealth Assoc., 1982) for the ultimate configuration of three single-circuit/-~45-kV lines in parallel alignment on common right-of-way (Fig. 0-2) and operating at maximum rated voltage of 362.5 kV. The results, calculated by computer program by CAl using algorithms described by Electric Power Research Institute (1975a) and given in Table D-8, agree closely with manual calculations performed by FfRC using formulas and design curves given in Electric Power Research Institute (1982). The calculated values apply for a 1 MHz simulated noise measuring frequency, which is centrally located in the AM broadcast band.'" Based ·on a 25 dB~ average level of measured ambient radio noise, this table indicates that clearances of 200, 300,and 600 ft (60, 90, and 180 m) from the edge of the right-of-way would probably be required under fair weather, wet conductor, or heavy rain conditions, respectively, to avoid the possibility of significant adverse impact on AM radio reception due to corona-generated radio noise. This is summarized in Table 0-9. Based on these calculations and survey measurements, it appears reasonably conservative to allow for a 1,000-ft (300-m) minimum separation between the corridor centerline and residences and between the centerline and long parallels with the Parks Highway, as recommended by APA/CAI."'''' This should be adequate to protect against significant radio noise contributions from the lines on AM broadcast reception under worst-case weather conditions and should ensure that existing quality of AM radio reception is preserved. However, mitigative measures can be taken to restore AM reception quality should particularly troublesome problems arise after line installation, such as installing a separate receiving antenna beyond the influence zone of the radio noise and connecting it to the receiver by a shielded lead-in cable. This or other means could be resorted to on a case-by-case basis within practical limits. Any problems that conceivably could be traceable to project transmission line operation would have to lie within areas where intelligible reception now exists. For the Willow-Healy section of the Anchorage-Fairbanks transmission corridor, this would include only the portion between Willow and Curry, since, as Table 0-7 shows, useful AM radio reception north of Curry is practically nonexistent. In interpreting the resul ts of thi s i nvesti gati on conducted by APA and its consul tants, it should be noted that the radio reception analysis appl ies only to AM radio reception under daytime conditions, i. e., by ground wave transmission. At night, the radiowave propagation characteristics change markedly, which can result in deterioration of reception quality. This influence and the static noise generated from thunderstorm activity can mask the effects of any line-generated noise. Radio interference to citizens band (CB) communications from a transmission lioe can come from two sources: (1) from a static-type interference caused by line corona or spark discharges from loose or broken line hardware, which can often be located and repaired; and (2) from the blocking action due to the physical presence of the line itself. The corona noise only causes receiving interference. The signal-to-noise ratio with regard to corona noise is likely to be quite high, since corona-generated radio noise falls off to negligible intensity beyond about 10 MHz, well below the 27 MHz of the CB band, in which region the noise will be about 250 times lower (48 dB~) than at 1 MHz, the mid-point of the AM broadcast band. Furthermore, atmospheric static will likely mask the radio noise at the time when the latter will be at its maximum (during thunder­ storms). 80th the sending and receiving modes could be affected by the blocking. However, since CB units are often mobile or portable, they can be easily moved to a location 100 ft (30 m) or more from the line, which should be sufficient to restore good CB communications. Therefore, CB interference from line-generated radio noise would not be anticipated to present a problem of any consequence. Television interference (TVI) can occur in the AM video portion of the television signal from radio noise."''''''' As in the case with radio interference (Rl), TVl generation on transmission

"'Values at the low frequency end of the broadcast band (550 kHz) are typically about 4.0 dB greater than the value at 1 MHz, and those at the higher end (1600 kHz) are about 5 dB lower, reflecting the dropoff with increasing frequency noted earlier (Electric Power Research Inst., 1982). "''''This 1,000-ft (300-m) separation is equivalent to 800 ft (240 m) from the edge of the 400-ft (120-m) right-of-way. "'''''''As previously mentioned, the audio portion of the television signal is broadcast on FM and is not subject to static interference.

~------0-16

Table D-8. CaJeulated Transmission Line Radio Frequency (RF) Noise Levels (three single-circuit transmission lines on a common right"'of-wayt! simulated noise measuring frequency = 1 MHz)

. Lateral Separation Weather Conditions from Edge of Heavy Rai nt2 Wet Conductort3 Fair Weather Right-of-Way (ft) (dBIJ) (dBIJ ) (dBIJ) 0 69 57 49 100 50 38 30 200 40 28 20t4 300 34 22t4 14t4 400 29 19t4 9t4 500 .' 25 15t4 5t4 600 22t4 12t4 2t4 700 20t4 10t4 Ot 4 800 18t4 8t4 -2t4 t 1 Configuration along part of the 345 kV Anchorage-Fairbanks trans- mission corridor. t 2 "Heavy rain" is considered as a natural rainfall rate on the order of 0.31-0.47 in/hour (8-12 mm/hr) or greater. It is the highest corona and radio-noise-producing condition. For 99% of the total foul-weather period, the radio noise generated by corona can be expected to be below the heavy rain value (Electric Power Research Instit., 1982). t 3 The "wet conductor" condition represents a natural condition of very light rain, drizzle, or dense fog when the conductor is satu-· rated with pendant water drpps and the concentration of moisture in the air is just sufficient to maintain an equilibrium between the loss and replacement of water drops (Electric Power Research Inst., 1982). t 4 Average value of measured ambient RF noise level during the Intertie route survey was about 25 dBIJ. Therefore, the corona-generated transmi-ssion line RF noise component is not expected to have a sig­ nificant impact on the existing quality of radio reception for the noted calculated levels corresponding to the indicated weather conditions and lateral distances from the edge of the right-of-way. Conversion: To convert feet to meters, multiply by 0.305. Source: Commonwealth Associates (1982). 0-17

Table 0-9. Zones of Influence of Radio Frequency Noise (three single-circuit transmission lines on a common right-of-wayt1 operating~t 362.5 kV)

Width of Distance from Zonet 2 Edge of Right-of-Wayf2;< 3 Weather Condition (ft) (ft)t " Fair weather 800 200 Wet conductort4 1000 300 Heavy raint4 1600 600 t 1 Configuration along part of the 345-kVAnchorage-Fairbanks transmission corridor. t 2 The right-of-way width is assumed constant at 400 ft. t 3 At greater distances from the edge of the right-of-way, no impact on the quality of radio reception from the operation of the Intertie is foreseen for the weather conditions cited. t 4 Defined in Table 0-8. Conversion: To convert feet to meters, multiply by 0.305. Source: . Commonwealth Associates (1982).

lines results from corona formation on the conductors as well as from spark discharges on loose or broken line hardware. TVI in the form of "ghost" images can also result due to the physical presence of the transmission line, causing signal reflections. Since corona-generated radio frequency noise drops off at the higher frequency levels and becomes quite low in intensity above 10 MHz, the TVI, even in the low television band (Channels 2-6, 54-88 MHz), will generally be negligible, particularly if the AM radio reception in the vicinity of the transmission line is acceptable (Commonwealth Assoc., 1982), as is expected to be the case for the corridor lines. Problems resulting from spark discharges on line hardware can generally be attended to satis­ factorily on a case-by-case basis by suitable repairs or local minor design modifications. Television signal reflection problems caused by transmission lines traversing sparsely settled rural areas should likewise be few in number and can be relieved by modification or relocation of the antenna on the receiving apparatus. A preconstruction television reception survey was made by APA/CAI similar to the radio reception survey described above in connection with AM radio reception (Commonwealth Assoc., 1982). The same 11 measuring sites were chosen along the corridor. Thirteen television signals were monitored, as given in Table 0-10. Of these, nine were from television translators, which are commonly used among the rural communities that 1ie along the corridor route. Translators are low-power facilities (normally 10 watts) that receive weak signals from primary television stations located in Anchorage and Fairbanks and rebroadcast the video and audio on a different channe1 to 1oca 1 small geographi ca 1 area withi n a 20- to 30-mi (30- to 50- km) radi us. Because the rebroadcasted signal is relatively much stronger than the weak primary signal, it is less susceptible to interference from the transmission lines. Quasi-peak measurements were made of the signal strength of the television signals that could be received at each of the 11 sites along with measurements of the ambient noise level at a clear frequency slightly below the video carrier frequency. The resulting signal-to-noise ratios ranged from 2 to 44 dB~, with an average of 12 dB~. Only two of the 34 values of SNR were above 30 dB~, the minimum level judged accept­ able for viewing by a 500-person test group in a study sponsored by the Electric Power Research Institute (1982). However, the qual ity of audio reception (FM), as judged by the observers making these measurements, was entirely satisfactory (A-rating) in some cases, as shown in Table 0-11. This is the same rating scale used in the evaluation of AM radio reception (Table 0.7). In summary, it appears that no significant television reception problems would be likely to develop as a consequence of the installation of the corridor transmission link, and that accept­ able TV reception should be preserved where present reception is good. Insofar as the remaining sections of the project 345-kV transmission are concerned, the results of the analytical study should still be valid [the recommended 1,000-ft (300-m) separation between the lines and residences and between the lines and long parallels with highways, etc.]. In fact, this requirement should be increasingly conservative for rights-of-way containing less .. than three circuits, although the effect of multiple circuit right-of-way occupancy is secondary 0-18

Table D-10. TV Stations Received During Preconstruction Survey of Corridor Route. July 1981

Operating Antenna Call Power (kW). Height (ft), Channel Letters Location Visual/Aural AT/AGt l 2· KEN I Anchorage 26.9/2.69 70/173 2 KFAR Fairbanks 5.37/.676 45/200 Cantwell translator at earth station operated by Alaska Department of Highways 4 K04CO Healy translator (Primary Ch. 11, KTVF Fairbanks) 4 K04DO Talkeetna translator (Primary Ch. 11, Anchorage) 6 K06KG Talkeetna Translator (Primary Ch. 13, Anchorage) 7t2 KAKM Anchorage 105/20.90 143/250 7 K0900 Healy translator (Primary Ch. 9. Fairbanks) 9 KUAC Fairbanks 46.7/1.16 200/255 9 K0900 Talkeetna translator (Primary Ch. 2. Anchorage) 11 KTVA Anchorage 26.3/5.35 300/391 13 KIMO Anchorage 30/6.17 90/347 13 Healy translator t 1 AT--above average terrain; AG--above ground. t 2 Non-commercial educational station. Conversion: ,To convert feet to meters, multiply by 0.305. Source: Commonwealth Associates (1982).

Table 0-11. Existing Quality of Television Reception (audio)t 1

Number of TV Signals Judged to Have the Following Quality of 2 Site Audio Receptiont Number Location A BC 0 E 10 Wi 11 ow 3 1 1 20 Trapper Creek 1 1 1 1 1 30 Chase 2 1 3 1 40 Lane Creek 2 3 50A Curry 1 60 Cantwell 1 1 1 70 Carlo Creek 1 80 Deneki Lake 2 90 McK in1ey Vill age 1 100 McKinley Park 119A Healy 2 1 1 Total 11 4 2 9 8 t 1 Based on field measurements of radio station signal strengths July 9-15, 1981...... t 2 A--entirely satisfactory; B--very good, background unobtrusive; C--fairly satisfactory, background plainly evident; D~-background very evident. speech understandable with concentration; E--speech unintelligible. Sou~ce: Commonwealth Associates (1982). D-19

with regard to radio noise generation. No surveys of radio or television broadcast reception qual ity have yet been reported by APAa long other transmi ssi on routes, although there seems to be no reason why reception of acceptable quality should not be preserved in locations where present reception is good, particularly with the 1,000-ft (300-m) recommended separation distance between the lines and receiving antennas.

A survey of existing potentially sensitive communication tower locations along th~ proposed corridor route was conducted by APA's consu1tants (Commonwealth Assoc., 1982) to form the basis of determining minimum separation distances between the edge of the- 345-kV transmission right­ of-way and these towers to preclude the likelihood of static noise or reflective type inter­ ference to the operation o~ these facilities due to corona formation on the conductors and the physical presence of the line conductors and towers. An additional objective was to ensure adequate safety clearance between the towers and the lines in the event of wind toppling and for tower maintenance. Necessary corrective measures could then be carried out prior to construc­ tion to minimize these impacts. Altogether, 50 such radio communication facilities were identi­ fied along the Anchorage-Fairbanks transmission corridor between Willow and Healy (Commonwealth Assoc., 1982: Fig. B-1). Included were FM translators, TV translators, earth stations (for communications with geostationary satellites), air navigational aids, and point-to-point micro­ wave facilities. These facilities are licensed to various business and governmental agencies in Alaska, including the Federal Aviation Administration, the Alaska Railroad, the Alaska Depart­ ment of Highways, the Golden Valley Electric Association, the Mantanuska Telephone Association, and Alascom, Inc. The criteria for minimum clearances for the various types of communication facilities is given in Table D-12. The basis for these criteria take into account both the operational interference and the safety aspects, discussed above. A survey of the owners of these facilities indicated, in general, agreement as to the guidelines outlined in the criteria, and no problems are antici­ pated to result from their use. The precise locations of the communication towers with respect to the corridor will be verified by aerial photographs. Surveys of communication facility locations along 345-kV transmission routes other than between Willow and Healy along the corridor have yet to be made by APA. However, the foregoing criteria should be applicable to these lines, also. D.5 ELECTRIC AND MAGNETIC FIELDS D.5.1 Electric Fields Energized overhead transmission lines generate electrostatic fields in the surrounding insulating medium that give rise to induced electric charges in other conducting objects in the vicinity of the line, such as vehicles, fences, rain gutters, etc. If these objects are insulated, or semi - i nsul ated from ground, the induced charge accumul ati on on the object produces a potential difference, or voltage, between the object and ground. Depending on the resistivity of the path between the object and ground, this induced voltage will give rise to a current flow, generally measured in milliamperes. If a person were to approach and touch this object (say, a parked vehicle), a second parallel path to ground would be created and a small current would pass through his body, the amount depending on his internal body resistance and the resistance from his body to ground. This might be accompanied by a mild shock and spark discharge. This mainly constitutes an annoyance and never (or seldom) results in physical injury, except due to a secondary reaction, such as an involuntary muscle contraction. The foregoing effects vary with the intensity of the electrostatic field, which is generally _ measured in terms of the field gradient (in kV/m) at "ground level" (or, more precisely, the gradient measured or calculated at 1 m above ground). This standardization permits comparisons to be made between different line designs in this respect. The earth's ambient DC electric field at ground level is 0.13 kV/m, although beneath thunder clouds the fields may reach 3 kV/m even in the absence of lightning (Bonneville Power Admin., 1977). The ground-level gradient is a valid parameter for the prediction of electrostatic effects and, for given conductor-to-ground heights, does not vary more than 10% to 15% for heights up to 10 ft (3 m) (Bonneville Power Admin., 1977). Electric fields under energized lines can be measured with instrumentation, but they also can be accurately calculated, enabling transmission lines to be designed with known field strengths at ground level. For a given line design and measuring point, the field strength varies directly with line voltage. It also varies with conductor bundle geometry, phase spacing, height of the conductors above ground, and lateral distance of the measuring' point from the line centerline. As this lateral distance increases, the field strength decreases, dropping rapidly and leveling off at a low value beyond the edge of the right-of-way. Therefore, the effects of the field directly under or very close to the line are of primary interest. For Anchorage-Fairbanks transmission corridor sections having three 345-kV lines on a common 400 ft (120 m) right-of-way and at maximum operating voltage of 362.5 kV, the maximum calculated .. · Tabl e 0-12. Possible EHV Line Effects on Communications Facilities and Recommended Minimum Clearancest1

Recommended Minimum Communication .- No Reported Clearance to Facil ity Reflection Diffraction Absorption Ghosting Effects EHV Lines Criterion

FM Translator X Antenna height Antenna Toppling Guy plus 200 ft Anchor Maintenance TV Translator X X X Antenna height Antenna Toppling Guy plus 200 ft Anchor Maintenance Earth Stations 10 tower heights Line of sight for low zenith of approximately 19° above the horizon NAVAJOS (Enroule) NDB X 1000 ft FAA RCAG X 1000 ft FAA SFO X 1000 ft FAA SSFO X 1000 ft FAA CI, N NAVAJOS 0 (At Airports) 1. 5° DOT/FAA VOR X 1. 5° FAA (1968) Unicorn X Airport Criterion RCO X Airport Criterion FSS X Airport Criterion AAS X Airport Criterion ALAS X Airport Criterion Point-to-Point X X X Antenna height Antenna Toppling and Mi crowave plus 200 ft Guy Anchor Maintenance 0.6 First Fresnel Zone Radius t 1 "X" denotes potential exists for specified effect. Conversion: To convert feet to meters, multiply by 0.305. Source: Commonwealth Associates (1982). '" D-21

electric field strength at ground level is 6.9 kV/m,* which occurs within the right-of-way limits near the outside phase conductors. At the edge of the right-of-way, the calculated field is-1.6 kV/m (Commonwealth Assoc., 1982). Other values as"-a function of distance from the right­ of-way centerline are given in Table 0-13, which shows the rapid dropoff of field intensity beyond the edge of the right-of-way. Calculations made for a large vehicle [13.5 ft high, $.5 ft wide, and 70 ft long** (4.1 x 1.6 x 21 m)J parked in this field and oriented transversely to the line indicated that a current of approximately 4.5 milliampere (rna) would.

*This is a computer-calculated value given in the Report R-2394 (Commonwealth Assoc., 1982), which is in reasonably close agreement and is conservative with respect to the 6. 74-kV/m value computed manually by FERC using methods described by the Electric Power Research Institute (1982). The 6.9-kV/m value is also below the 8.0-kV/m limit set in the Anchorage-Fairbanks Intertie Basic Design Criteria (Commonwealth Assoc., 1981). **The largest size permitted on Alaskan State highways.

-1IiIIlIl rene, ttm' ~1fr@~..:~;CJ:.::..'.:.:;;,L;,.;;..-~;~,i'~;':;'ili';f~;,;;&:i*N'#i""·''h'j;; '"".:iW"t¥i"i D-22

Table D-13. Calculated Intertie Electric Field Strengths (three transmission lines operating at 362.5 kV)tl

Lateral Separation Calculated Electric Field from Centerline of Strengths (kV/m). Right-of-Way (ft) at 1 m above Ground Levelt2 o 5.14 50 3.78 100 3.90 150 6.90t3 200 1. 6lt4 300 0.22 400 0.07 t 1 Configuration along part of the Anchorage-Fairbanks 345-kV transmission corridor. t 2 Values are calculated for a minimum conductor height (atmid~span) of 30 ft and a 33-ft phase-to-phase spacing. t 3 The maximum value of electric field strength occurs near the outside phase conductors. t 4 Value of electric field strength at th~ edge of the right-of-way. . Conversion: To convert feet to meters, multiply by 0.305. Source: Commonwealth Associates (1982).

1982). Si nce thisis al so one to two orders of magnitude smaller then the magnetic fields produced by common household appliances (Bridges, 1975), its effect is considered harmless. Unbalanced phase loadihg can produce induced voltages up to ten times those of the balanced condition on account of the absence of cancellation in magnetic field coupling caused by the . dissymmetry (Electric Power Research Inst., 1975b). However, under normal conditions the phase loadings are nearly perfectly balanced. Significant unbalances are likely to result only during fault conditions and then only for extremely short periods of time necessary for fault clearing. In the balanced loading case, the induced current density in a person standing close to the line from the magnetic field is on the order of 1/10 that resulting from the electric field induction. Furthermore, the magnetic fields do not Cause transient currents of high peak value, which can result from electric-field-induced spark discharges (Electric Power Research Inst., 1975b). As in the case of electric fields, experimental findings with regard to biological effects from continued exposure to magnetic fields are inconclusive (Sheppard and Eisenbud, 1977; Electric Power Research Inst., 1982). Based on a survey of the literature reported by the Electric Power Research Institute (1982), the inconclusive nature of the results is due in part to the wide variatioriin field strengths, frequencies, and exposure durations used in different studies, One such study noted an increase in triglyceride in humans exposed to a I-gauss field of 45 Hz for 22.5 hours. However, these findings were not duplicated in more controlled tests on monkeys in fields up to 10-gauss at 45 Hz. Another study in Germany showed no effect on reaction times, pulse frequency, arterial pressure, or electrocardiographic and electroencephalographic traces on three volunteers sUbjected to a 3-gauss field at 50 Hz. These findings, which generally agree with those from studies of the effects of electrostatic fields, have resulted in little concern ,for the~biological effects of magnetic fields, particularly in view of the relatively lower level of lnduction from magnetic fields than from electric fields. D.6 ELECTRICAL SAFETY Physical contact with energized transmission line conductors, either directly or through other inetallic objects will, of course, be fethal, the same as is likely with accidental contact with residential distribution lines, which operate at much lower voltages. Precautions should there­ fore be exercised when operating or transporting under the line any kind of apparatus or equip­ ment that exceeds normal vehicle height to prevent accidental contact with the line conductors, either directly or by flashing. D-23

For the proposed project 345-kV transmission lines, the minimum vertical line clearance of phase conductors above ground at 120 0 F (49°C) would be 30 ft (9 m) for farmland and highways, which exceeds the National Electric Safety Code (NESC) minimum or 27.3 ft (8.3 m) (American National Standards Inst., 1984). This and other dimensions pertinent to safety clearances and right-of­ way occupancy are given in Figure 0-2 for the Knik Arm-Gold Creek section of the Anchorage­ Fa irbanks 345- kV transmi ssi on corridor. Ri ght-of-way wi dths and spaci ng of 1i nes ~pp 1icab 1e to right-of-way occupancies of one to four single-circuit lines are given in Table D-:tLt-;:

Table D-14. Right-of-Way Use of Single and Multiple Single-Circuit Transmission Lines

No. of Single- Width of Lateral Separation Distance from Line L Circuit Lines ROW (ft) of Line L (ft) to Edge of ROW (ft) 1 190 -- 95 2 300 105± 95 3 400 105± 95 4 510 105± 95 Conversion: To convert feet to meters, multiply by 0.305. Source: Application ExhibitF, Plate F81.

REFERENCES FOR APPENDIX D American National Standards Institute. 1984. National Electrical Safety Code. 1984 Ed., C2. Institute of Electrical and Electronics Engineers, Inc. New York. Anonymous. 1974. Sound and Vibration Magazine. pp. 33-36 (September). Bonneville Power Administration. 1977. The Role of the Bonneville Power Administration in the Pacific Northwest Power Supply System. Appendix B. BPA Power Transmission. Bridges, J.E. 1975. Biological Effects of 60 Hertz Electric Fields. ITT Research Institute. Final Report E8151, EPRI Project RP-381-1. Commonwealth Associates, Inc. 1981. Anchorage-Fairbanks Transmission Intertie, Basic Design Criteria. Prepared for Alaska Power Authority. Commonwealth Associates, Inc. 1982. Electrical Environmental Effects Report. Engineering Report R-2394. Prepared for Alaska Power Authority. Electric Power Research Institute. 1975a. Transmission Line Reference Book, 345 kV and Above, 1st Ed. Electric Power Research Institute. 1975b. Electrostatic and Electromagnetic Effects of Ultrahigh-Voltage Transmission Lines. Report EL-802, EPRI Project 566-1, Final Report. Electric Power Research Institute. 1982. Transmission Line Reference Book, 345 kV and Above, 2nd Ed. Federal Aviation Administration. 1968. VOR/VORTAC Criteria, 6700.11. Federal Communications Commission. 1968. Federal Communication Commission Rules and Regulations. Vol. III. Federal Communications Commission. 1975. Federal Communication Commission Rules and Regulations. Vol. II, Part 15. Revised to May 1975. Fern, W.J. and R.I. Brabets. 1974. Field investigation of ozone adjacent to high voltage transmission lines. Institute of Electrical and Electronic Engineers. Transactions on Power Apparatus and Systems, Vol. PAS-93(5). Frydman, M., A. Levy and S.E. Miller. 1972. Oxidant measurements in the vicinity of energized 765 kV lines. Institute of Electrical and Electronic Engineers. Transactions on Power Apparatus and Systems, Vol. PAS-92(3).

-d ~,- 0-24

Frydman, M. and C. H. Shih. 1973. Effects of the environment on oxidants production in AC corona. Institute of Electrical and Electronic Engineers. Transactions on Power Apparatus and Systems, Vol. PAS-93(1). Institute of Electrical and Electronic Engineers. 1979. The Electrostatic and Electromagnetic Effects of AC Transmission Lines. Tutorial Course 79 EH0145-3-PWR. New York. Mahmoud, A.A. and D. Zimmerman. 1982. Pigs fare well under 345 kV line. Transmission and Distribution Magazine (December). National Academy of Sciences. 1977. Guidelines for Preparing Environmental Impact Statements ~n Noise. Committee on Hearing, Bioacoustics, and Biomechanics. Washington, DC. Riog, R.. 1979. The effects of transmission Lines. Record of the Maryland Power Plant Siting Act~ Vol. 7, No.1. Scherer, H.N., B.J. Ware and C.H. Shih. 1972. Gaseous effluents due to EHV transmission line corona. Institute of Electrical and Electronic Engineers. Transactions on Power Apparatus and Systems, Vol. PAS-92 (3). Sebo, S.A., J.T. Heibel, M. Frydman and C.H. Shih. 1975. Examination of ozone emanations from EHV transmission line corona discharges. Institute of Electrical and Electronic Engineers. Transactions on Power Apparatus and Systems, Vol. PAS-95(2); Sheppard, A.R. and M. Eisenbud. 1977. Biological Effects of Electric and Magnetic Fields of Extremely Low Frequency. New York University Press. New York. U.S. Environmental Protection Agency. 1974. Information on Levels of Environmental Noise Requisite to Protect Public Health and Welfare with an Adequate Margin of Safety. NTIS PB-239429.